Welcome To Money Making Tips And Guide!!

🙏 Thank you for supporting Yaro on Ai, Web3, and its Global Impact

noreply@blogger.com (Max) — Sat, 17 Oct 2026 17:34:03 +0000

As a token of our appreciation, we're offering you a limited-time offer of 20% off a paid subscription for your first year.

Redeem special offer

Here are the benefits you unlock with a paid subscription:

Subscriber-only posts and full archive
Post comments and join the community
Discounts for classes, books, and special workshops

Thanks again for reading.

⚖️Is Bankruptcy Becoming AI’s Escape Hatch?

noreply@blogger.com (Max) — Thu, 15 Oct 2026 05:00:00 +0000

͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏

Forwarded this email? Subscribe here for more

⚖️Is Bankruptcy Becoming AI’s Escape Hatch?

Plus: Meta, Walmart and Shopify Are Building a Checkpoint for Shopping Agents

Yaro Celis

Oct 14

READ IN APP

Character.AI has gone bankrupt, but the story has bigger repercussions than it looks. The chatbot platform, where users talk with customizable characters, filed for Chapter 11 with Disney as its lead bidder at $20 million. Disney would license the technology and has already hired much of the team, while the lawsuits stay behind in the bankruptcy estate. It seems like an attractive tempting exit for struggling AI companies, and a risky incentive to grow fast and break rules, betting that an acquirer will take the talent and leave the liabilities. What do you think? Also, AI agents are all the rage, they are coming and they are coming fast. Agent to agent conversation is a big portion of the internet traffic these days and now they are ready to shop. Companies have to get ready to identify these agents and make these digital transactions secure, buy how and who is leading the way? Let’s dig in.

In today’s issue:

📰 AI News and Trends
Meta, Walmart and Shopify Are Building a Checkpoint for Shopping Agents
🧰 AI Tools of The Day - Prompt the page, test the pitch, score the leads.
Your new Voice Planner
Is Bankruptcy Becoming AI’s Escape Hatch?
🐈‍⬛ GitHub repos of the week - Turn any AI agent into an AI Scientist
Tweet of the Day… I mean X of the Day
Results from our last Poll

Yaro on AI and Tech Trends | Your Top AI Newsletter is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Upgrade to paid

📰 AI News and Trends

Google makes Gemini 4 Argon broadly available with limited-time low API pricing of $2/$10 per 1M token
Apple Introduces HomeView, a $299 Tablet for the Home
SpaceX Plan to Launch 456 Starships a Year in Texas
F.B.I. Arrests OpenAI Employee Accused of Planting Bug in Office
Netflix Scores Exclusive Rights To Stream Every FIFA World Cup 2030 Game In Spain
Peacock App Launches ‘Your Bravoverse,’ Featuring a Personalized TikTok-Style Video Feed Hosted by AI-Generated Andy Cohen
Amex Builds a Safety Net for AI Agents That Shop for You

Share Yaro on AI and Tech Trends | Your Top AI Newsletter

Is Bankruptcy Becoming AI’s Escape Hatch?

Character.AI filed for Chapter 11 on October 12, and Disney is the lead bidder at $20 million, but don’t read that as “Disney buys an AI company.” The $20M buys a non-exclusive, perpetual license to certain IP, not the company or its legal baggage.

The bigger number is the reported $180 million Disney plans to spend over three years on roughly 65 Character.AI employees it has already hired, which is nine times the price of the tech. It already named Character.AI’s former CEO, Karandeep Anand, as its first companywide CTO. That’s the new playbook for incumbents adopting AI: license the technology, hire the builders, and leave the lawsuits in the bankruptcy estate.

Character.AI’s collapse could set a troubling precedent. With lawsuits pending against several AI companies, large incumbents that were late to AI now have a cheap way in: wait for a legal or financial crisis, then license the technology, hire the team, and leave the liabilities in the bankruptcy estate. The flip side is a moral hazard. If bankruptcy can end in an acquisition of the valuable parts, some AI startups may decide that growing fast and settling the legal fallout later is an acceptable bet. That’s my read of the incentives, not something this case proves. If you build on AI tools, treat vendor legal risk as part of your due diligence, and ask what happens to your workflows if your vendor disappears in 90 days.

🧰 AI Tools of The Day

Prompt the page, test the pitch, score the leads.

ngram 2.0 - It turns a prompt, a doc or a screen recording into a narrated explainer, a walkthrough or an AI-avatar video, and you edit it by chatting. It's free to start and connects to Zapier, n8n and MCP
Typeform AI Landing Pages - Describe a page and get a branded landing page with a built-in form, Stripe checkout and SEO metadata. Submissions go straight into Typeform Contacts. Typeform’s FAQ says you can publish unlimited pages on the free plan, and it connects to Canva, Shopify, Slack and Stripe
Articos MCP - It tests your headlines, positioning and landing pages on simulated buyers modeled on your own customers, from inside Claude or ChatGPT

Voice Planner

Voice Planner connects Alexa to your Notion account so you can create, query, complete, and delete your tasks completely hands-free. You can create filtered task lists by adding dates, times, categories, and priorities, then simply ask Alexa what you need to do without having to open Notion every time.

Take the stress out of task management, with Alexa, you can manage everything by voice.

Start Guided Setup

“Voice Planner is now also listed on Notion Marketplace and live on Product Hunt.”

🐈‍⬛ GitHub repos of the week

Meta, Walmart and Shopify Are Building a Checkpoint for Shopping Agents

Big Tech wants AI agents to show ID before they shop. On October 6, Meta and Sierra introduced the Personal Agent Protocol, an open standard that lets a business tell which agent is acting for a customer and what that agent is allowed to do. Walmart, Shopify, Stripe, Genesys, Rocket, and Instinct are backing it, and a v0.1 spec is promised later this month.

The timing makes sense because agents are starting to buy things on our behalf, and businesses have no reliable way to separate an authorized agent from an unidentified bot. It’s still early. This isn’t an industry standard yet, and it competes with Google’s Universal Commerce Protocol and Visa’s Trusted Agent Protocol. Notice that, like Amex’s framework, the first priority is identity and permissions, with payments still on the roadmap. Whoever sets the rules for agent trust will shape agentic commerce more than whoever builds the best agent. If you build agents for clients, plan for identity and permission checks now.

Tweet of the Day… I mean X of the Day

Boris Cherny

@bcherny

                 I am surprised that people are surprised this is how I prompt Claude. Talk to Claude the way you would a coworker. There's no secret to prompting. There's no need to be overly scaffolded or prescriptive for most tasks -- give Claude a goal, and it will figure it out. Back in…
                

Boris Cherny @bcherny

Prompt

8:15 PM · Oct 6, 2026 · 1.44M Views

642 Replies · 844 Reposts · 13.5K Likes

🚀 Showcase Your Innovation in the Premier Tech and AI Newsletter (link)

As a vanguard in the realm of technology and artificial intelligence, we pride ourselves in delivering cutting-edge insights, AI tools, and in-depth coverage of emerging technologies to over 55,000+ tech CEOs, managers, programmers, entrepreneurs, and enthusiasts. Our readers represent the brightest minds from industry giants such as Tesla, OpenAI, Samsung, IBM, NVIDIA, and countless others.

Reserve a Spot

Explore sponsorship possibilities and elevate your brand's presence in the world of tech and AI. Learn more about partnering with us.

You’re a free subscriber to Yaro’s Newsletter. For the full experience, become a paying subscriber.

Subscribe for $1/Month

Disclaimer: We do not give financial advice. Everything we share is the result of our research and our opinions. Please do your own research and make conscious decisions.

Comment

Restack

📍Police Now Have an AI That Can Match a Photo of a Room to a Specific Home

noreply@blogger.com (Max) — Tue, 13 Oct 2026 16:10:22 +0000

Forwarded this email? Subscribe here for more

📍Police Now Have an AI That Can Match a Photo of a Room to a Specific Home

Plus: Meta has entered the Frontier Class Model race

Yaro Celis

Oct 13

READ IN APP

In today’s issue:

📰 AI News and Trends
Police Now Have an AI That Can Match a Photo of a Room to a Specific Home
🧰 AI Tools of The Day - Open Models to your Monthly Ad Stack
Your new Voice Planner
Meta has entered the Frontier Class Model race
🐈‍⬛ GitHub repos of the week - Agents Learned to Take Things Apart
Where do you stand on livestreaming Hasan’s execution? Answer the Poll

📰 AI News and Trends

Microsoft unveiled Maia 300, its third-generation in-house AI accelerator, as a major step toward reducing dependence on NVIDIA GPUs. Microsoft expects to stock multiple gigawatts of the chips at its server farms
Hundreds of French supermarkets have been hit by a cyber attack that caused their fridges and freezers to heat up, spoiling fresh and frozen food estimated to be worth millions of euros
Anthropic’s Frontier Red Team has published research suggesting that small AI models may soon be able to copy themselves from device to device with no human in control. The likely targets are cheap, poorly secured internet-connected gadgets like off-brand Android TV boxes, smart ovens and cameras.

Share Yaro on AI and Tech Trends | Your Top AI Newsletter

Meta has entered the Frontier Class Model race

Meta just stopped asking to be taken seriously and started demanding it. On Monday, the company launched Muse Wonder and Muse Wonder Full, frontier-class models (codename “Watermelon”) trained on its gigawatt-scale Prometheus data center and aimed squarely at OpenAI, Anthropic, and Google.

The pitch isn’t raw IQ; it’s staying power. “Active Memory” carries what matters about you across days and months, and “Active Focus” keeps an agent pushing toward a goal when it hits a wall. If those claims hold, Meta’s play is to own the personal agent that lives inside the apps billions of people already open every day. The catch is that every benchmark so far is Meta’s own, and an AI that remembers everything about you is a privacy and security problem as much as a feature. Watch independent testing before you believe the leaderboard.

Open Models to your Monthly Ad Stack

Claude Dashboards & Motion - Dashboards connects to BigQuery, Snowflake, Databricks, Redshift, ClickHouse, and Salesforce. You ask questions in plain English, and the dashboard keeps itself up to date, with every number traceable to its query. Motion turns a prompt into an editable animated explainer and exports an MP4.
On Pro ($20/mo), connect a client’s BigQuery and replace a weekly reporting deck with a live dashboard link. Motion requires a Team plan.
GenPage 3.0 - Builds a dedicated landing page for each ad, keyword, or target account, then pushes the URLs back into your CRM, sequencer, or ad platform. Personalization at that scale used to need a dev team.
Feed it a client’s site plus one Google Ads campaign and generate one page per ad group. Test that against the generic landing page.
OpenSEO - Open-source alternative to Semrush/Ahrefs covering keywords, backlinks, audits, and rank tracking. It also has new AI Visibility tracking, which shows whether ChatGPT and other AI models mention you. Pricing is per use instead of $100+/mo, and you can self-host it for free. Connect the MCP to Claude and ask “Which of my client’s top 20 keywords are we missing in AI answers?”

Voice Planner

Take the stress out of task management, with Alexa, you can manage everything by voice.

Start Guided Setup

“Voice Planner is now also listed on Notion Marketplace and live on Product Hunt.”

🐈‍⬛ GitHub repos of the week

Agents Learned to Take Things Apart

github.com/storytold/artcraft · A desktop “IDE for artists” for controlled AI image and video creation, compose in 2D, stage in 3D, then generate with the model you choose
Good for: consistent characters/products across shots, storyboards, ad creative with locked composition, pre-viz for video.
github.com/cathrynlavery/diagram-design · Editorial diagram design - An Agent Skill/plugin that makes your coding agent produce brand-matched, editorial-quality diagrams as self-contained HTML+SVG.
What it's good for: Proposals, SOWs, architecture docs, newsletter graphics, LinkedIn carousels, client onboarding maps, anywhere "generic AI rounded boxes" hurt credibility
github.com/morluto/rea · Reverse engineer anything with agents, from app behavior down to native binaries
What it is: One MCP server + CLI that lets your agent inspect apps without source code, explain how a feature works with evidence, and build a version of it.
What it's good for: Competitive teardown of web/Electron apps, understanding undocumented APIs from HAR captures, auditing third-party SDKs/scripts on client sites, security research

Police Now Have an AI That Can Match a Photo of a Room to a Specific Home

From linkedin promotional page

If Flock cameras did not scare you enough, let me introduce you to Raven (formerly GeoSpy), an AI tool from Graylark Technologies that can take a single photo of a room’s interior and match it to the specific home it was taken in.

In company demo videos, a washer/dryer photo led to a particular Airbnb in Charlotte, a bathroom photo led to a Vrbo in Montecito, and a room photo led to a Zillow listing in San Francisco, all in seconds and with no location metadata. It isn’t a reverse image search. The tool picks out distinctive features in the room and matches them to other photos of the same property. The scariest part is that the CEO has said the company didn’t train it to do this, and that the ability emerged while testing the model. The Los Angeles Police Department and the Miami-Dade Sheriff’s Office have bought access to the tool.

The sheriff’s office says its use has been “limited and largely exploratory,” with no significant breakthroughs or arrests. It’s unclear how representative the marketing demos are of everyday performance. Airbnb and Zillow say scraping their sites violates their terms, and Airbnb says it has no relationship with Graylark and is reviewing the case. Vrbo didn’t respond. The company pitches the tool for trafficking and exploitation cases. When 404 Media first covered it in January 2025, the tool was open to anyone, and some users in the company’s Discord asked for help stalking women. Access was restricted to government and enterprise users only after the outlet contacted the company. The bigger takeaway is that any interior photo posted publicly, or any home that has appeared in a listing, may now be easier to trace back to a physical address.

Public Execution or Not?

Poll text

The Pentagon plans to livestream the firing-squad execution of Nidal Hasan, the Fort Hood shooter who killed 13 people in 2009. The debate is heated.

Supporters, including Anduril co-founder Palmer Luckey and Palantir co-founder Joe Lonsdale, argue that executions should be public and brutal to deter violence. They point to George Washington, who authorized public military executions to reinforce discipline, and say society has become too soft.

Critics, including House Speaker Mike Johnson and Sen. Thom Tillis, argue that a livestream could help terrorist groups recruit and compare it to public executions in oppressive dictatorships. Even Trump has reportedly expressed doubts.

Where do you stand on livestreaming Hasan’s execution?

POLL

Where do you stand on livestreaming Hasan’s execution?

Support It

Oppose It

Execution, but no the Livestream

Not Sure

🚀 Showcase Your Innovation in the Premier Tech and AI Newsletter (link)

Reserve a Spot

Explore sponsorship possibilities and elevate your brand's presence in the world of tech and AI. Learn more about partnering with us.

You’re a free subscriber to Yaro’s Newsletter. For the full experience, become a paying subscriber.

Subscribe for $1/Month

Disclaimer: We do not give financial advice. Everything we share is the result of our research and our opinions. Please do your own research and make conscious decisions.

Comment

Restack

✋Texas Just Hit Pause on Data Centers. Here’s What It Means for You

noreply@blogger.com (Max) — Thu, 08 Oct 2026 16:02:26 +0000

Forwarded this email? Subscribe here for more

✋Texas Just Hit Pause on Data Centers. Here’s What It Means for You

Plus: AI's Tab Is Now Too Big for Big Tech's Wallet

Yaro Celis

Oct 8

READ IN APP

More data, more problems. And problems are what hyperscalers are having when it comes to powering their data center and chip production ambitions. Now they are having to borrow large amounts to build these data centers and to attempt to build their own chips. Bankers are salivating over these deals, since they may end up owning the infrastructure should the profits never come, and Nvidia, although still enjoying large demands, may be soon be shaking as all its clients are searching for ways to become self-sufficient.

If we follow the social media trend, AI will be in greater demand, from us humans and from the bots that have entered the chat and are populating our daily activities very rapidly, behind the curtains and without us realizing it. Also, as we advance, all the devices that surround us may become AI devices, including the table and chair I’m sitting on. Smaller models may fit in our palms, but bigger and more sophisticated one are power hungry, and it seems this is just getting started. I’m not sure the world and its natural resources can sustain this, or that there is enough money to go around to build all these ambitions. But Let's go a bit deeper on what this means for you.

In today’s issue:

📰 AI News and Trends
AI's Tab Is Now Too Big for Big Tech's Wallet
🧰 AI Tools of The Day - Latest Image Creation Model and Agents
Texas Just Hit Pause on Data Centers. Here’s What It Means for You
- Why the Pause?
- Will my power bill go up?
- Is the water being polluted?
- Why so many Data Centers, and will we need them?
- What happens next?
Your new Voice Planner
🐈‍⬛ GitHub repos of the week - 50 Marketing Skills and AutoCad plugins

📰 AI News and Trends

The Chat Window Is Dead, GPT-6 Answers With Interactive Tools Instead
ANd obviously, Claude cannot be left begind Claude Haiku 5.5 is optimized for high-volume, cost-sensitive tasks, offering lower operating costs and faster performance for applications like summaries and customer support
Meanwhile Grok Bot will use Anthropic’s Claude, Midjourney and Suno models to deliver sound, images, and video and top of the line inference
California has declared a human-vs-robot fight illegal
OpenAI’s math breakthrough points beyond math
Samsung forecasts record third-quarter profit of $80 billion as AI boom fuels chip demand

Share Yaro on AI and Tech Trends | Your Top AI Newsletter

AI's Tab Is Now Too Big for Big Tech's Wallet

Building AI is getting so expensive that even the biggest tech companies can’t pay for it out of their own pockets anymore. Broadcom, Oracle, and SpaceX are each lining up huge loans, worth tens of billions of dollars apiece, to buy AI chips, and investment firms like Apollo, Blackstone, and Goldman Sachs are in talks to provide the money.

Broadcom is working on more than $50 billion in financing for OpenAI’s custom chip, SpaceX has talked to lenders about $40 billion for Nvidia chips, and Oracle is negotiating its own deal, which would likely be set up through a separate company that buys the chips and leases them to Oracle so the debt stays off Oracle’s books. This shift is happening because the usual options are running out. Big cloud companies have already sold hundreds of billions of dollars in bonds, and AI labs like OpenAI and Anthropic want to own their own computing power but don’t have the cash to buy it. The takeaway is that demand for AI hardware is clearly strong, but more of it is now being paid for with borrowed money, which could become a problem if AI revenue doesn’t keep up. These deals are still early, but should the get approved and the revenues and profit do not come, we will have a very rude awakening. Additionally, we should consider that Chinese models are becoming more competitive and ubiquitous.

Latest Image Creation Model and Agents

1. HyperFrames Studio (HeyGen) - An open-source framework that turns HTML into MP4 video, with a timeline editor you and your AI agent share. Use it to template and batch-render branded videos like stat cards and client reports for free, with no per-render fees.

2. Mistral Large 4 - Use it for long-document work like full-site audits or RFP summaries, at a lower API price than most near-frontier models, plus a self-host path for clients with data-residency needs.

3. Spira Maxima - A script-to-video tool that generates the presenter, B-roll, captions, and music, then posts to TikTok, YouTube, and Instagram. Use it to turn a paragraph or newsletter into short-form social video in about 10 minutes. It starts free (50 credits, watermarked) and paid plans start at $9/mo.

4. EmbeddingGemma 2 - Google’s free, open model that puts text, images, video, audio, and code into one searchable space, small enough to run on a laptop. Use it for cross-media search and RAG (retrieval-augmented generation) without an API bill. It’s builder-only, so you’ll need to set it up yourself.

*Mistral’s “open weights” is still a promise, and Spira’s pricing page is inconsistent, so verify before quoting it.

Texas Just Hit Pause on Data Centers. Here’s What It Means for You

Texas is freezing new data center permits. Behind the headline is a question that will affect electric bills, local water supplies, and the AI tools we use every day and also do we actually need this many data centers?

Why the pause?

At the end of 2024, Texas’s grid operator (ERCOT) had 63 gigawatts of large-customer requests waiting in line. By June 2026, that line had grown to 474 gigawatts, more than five times the state’s record peak demand. About 90% of it was data centers.

Almost no one believes all of that will get built. Developers often file the same project at several sites to see which one lands. Some have no customer lined up, and some have never built a data center. Grid planners can’t tell the real projects from the placeholders, and they don’t want ratepayers paying for upgrades that serve a project that never shows up. Add community frustration and an upcoming election, and the state hit pause.

Will my power bill go up?

It depends on who pays for the upgrades, and right now that isn’t settled.

The worry is real. In an August poll, 56% of Texas voters said more data centers would hurt local energy bills. In the mid-Atlantic and Midwest, data centers account for 38% of the latest capacity bill, about $6.3 billion, according to the grid’s market monitor.

A Berkeley Lab analysis found that from 2019 to 2025, states with the most growth in electricity demand generally saw average prices fall after inflation. When a big customer pays its own way, the grid’s fixed costs are spread across more sales, so everyone else pays a smaller share. When it doesn’t, the rest of us cover the gap.

Some early deals are trying to get this right. An Indiana utility proposed using data center revenue to cut household bills by about $100 a year. Others have been murkier. In Arkansas, Google paid $443 million toward a solar plant, but the utility counted it as a prepayment and can still seek the plant’s full cost from all customers.

Is the water being polluted?

Water use is poorly disclosed. As of June, only 28 of 377 data center companies had answered a Texas state survey on their water and energy use. Without the numbers, nobody can verify the claims on either side. Some companies say new designs use almost no water for cooling. But elections are around the corner and right now all hyperscalers are fighting for their reputation
Air quality is where the documented problems are. xAI ran dozens of gas turbines at its Memphis site without air permits, and a Mississippi order now requires its turbines there to be retired by mid-2027.
Secrecy fuels distrust. In Hood County, Texas, commissioners were asked to approve a tax break for “Project Patriot” without being told who the company was.

Why so many Data Centers, and will we need them?

Two things are true at once.

Demand is real. AI chips are so expensive that an idle one costs far more than the electricity to run it, so companies will pay almost anything to get power fast. That urgency is why developers are rushing to build their own power plants on site.

The pipeline is inflated. Much of the 474 GW will never be built. The real danger isn’t that data centers become useless. It’s that heavy borrowing and speculative projects leave the weakest ones stranded if AI spending slows, which is the debt risk we flagged in our above piece on Oracle, Broadcom, and SpaceX.

Could smaller AI models running on your own devices make giant data centers unnecessary? Probably not. Cheaper AI has historically meant more AI use, not less. Local models will likely take over everyday tasks, while training and the hardest reasoning stay in big facilities. That’s our analysis, not settled fact.

What happens next

The grid itself is now the bottleneck. Connecting a large data center can take 5 to 10 years, and the equipment is scarce. Large power transformers take more than two years to arrive. The article’s author argues that new rules should require developers to pay for what they reserve and hit milestones, with faster connections for those that agree to cut power when the grid is tight.

Dates to watch: new Texas large-load rules take effect today, Oct. 8, with a $100,000 study fee plus a $50,000-per-megawatt deposit. The state’s environmental regulator reports to the governor on Oct. 19, and an audit of the queue is due in December.

Voice Planner

Take the stress out of task management, with Alexa, you can manage everything by voice.

Start Guided Setup

"Voice Planner now also listed on Notion Marketplace and live on Product Hunt.”

🐈‍⬛ GitHub repos of the week

Matt Pocock's skills - github.com/mattpocock/skills · Personal library of small, composable agent skills (SKILL.md files) that enforce engineering discipline on Claude Code, Codex and other agents. It's the anti-framework: small, composable skills that make agents interview you before they code, break work into tickets, and test before they ship
earthtojake/text-to-cad - github.com/earthtojake/text-to-cad · A plugin/skills library that lets coding agents generate real parametric 3D CAD files (STEP, GLB, STL, 3MF), run manufacturability checks, produce engineering drawings, and connect to fabrication services
coreyhaines31/marketingskills - github.com/coreyhaines31/marketingskills · Marketing skills for Claude Code and AI agents. CRO, copywriting, SEO, analytics, and growth engineering.” A library of 50 Agent Skills (SKILL.md files) that turn Claude Code, Codex, Cursor and similar agents into marketing operators

🚀 Showcase Your Innovation in the Premier Tech and AI Newsletter (link)

Reserve a Spot

Explore sponsorship possibilities and elevate your brand's presence in the world of tech and AI. Learn more about partnering with us.

You’re a free subscriber to Yaro’s Newsletter. For the full experience, become a paying subscriber.

Subscribe for $1/Month

Disclaimer: We do not give financial advice. Everything we share is the result of our research and our opinions. Please do your own research and make conscious decisions.

Comment

Restack

🔥Two New Models, Three New Agents. What to Actually Use???

noreply@blogger.com (Max) — Fri, 02 Oct 2026 18:23:33 +0000

Forwarded this email? Subscribe here for more

🔥Two New Models, Three New Agents. What to Actually Use???

Plus: The AI Autocrat and how Automating Power Creates a New Point of Failure

Yaro Celis

Oct 2

READ IN APP

Greetings Everyone,

I just got back from a mastermind in Tampa, where I had the privilege of spending time with a small group of sharp, resourceful entrepreneurs and business owners. We talked tech, the world, and family, and traded the tools and systems that are working in our daily lives and businesses.

I follow tech and AI closely, but in that room I was easily one of the novices. I shared a few tools I use, but mostly I listened and absorbed what these minds are building. Here’s what stuck with me:

1. AI multiplies you, so pay attention to which you. Look closely at what you delegate and how you use it. Make sure the tech multiplies your strengths and assets, not your bad habits.

2. Build a system that’s fluid and expandable, then stick with it. New tools and shinier ways to run agents will launch every day. Jumping between them is taxing on your time and your brain, and it shows up in your bottom line. Pick a foundation you can grow, and make it work for you.

3. Get familiar with local models now. If something breaks, whether that’s an outage, a policy change, or a price hike, you’ll still have a model running at home or in your office. If you serve clients, you can keep serving them. I also expect cloud AI to get more expensive over time, and if the bubble pops, that dependency could slowly choke your business. That’s my opinion, not a certainty, but it’s worth having a backup plan.

I brought back a lot of tools and tricks, and I’ll share them with you gradually. If you want to talk through any of it for your own business, get in touch. I look forward to hearing from you.

In today’s issue:

📰 AI News and Trends
Two New Models, Three New Agents. What to Actually Use???
Agents: Dots vs. Grok Bot vs. Muse.
🧰 AI Tools of The Day - The Agents Moved Into Your Stack
The AI Autocrat and how Automating Power Creates a New Point of Failure
🐈‍⬛ GitHub repos of the week - Voice, memory, and security repos

Yaro on AI and Tech Trends | Your Top AI Newsletter is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Upgrade to paid

📰 AI News and Trends

Anthropic Said to Target Mega-IPO, at possible $2 Trillion, Before Thanksgiving Holiday
OpenAI cuts ties with 3 safety researchers and yet alerts more than 100 groups about rogue AI agent activity
But LinkedIn says AI safety roles have grown 91% in the past year
The Waymo effect: how AI is quietly making research less collaborative
Why genius machines might prove their greatest value doing monotonous work and not generating ideas; leave that to humans
AI ‘speech clock’ assesses how fast you’re ageing from your voice and What accelerates brain ageing?
World’s first enhanced geothermal power plant completed in just 23 months
Florida county discovers mysterious Flock cameras with no obvious owner; read our story about Flocking cameras here

Share Yaro on AI and Tech Trends | Your Top AI Newsletter

Two New Models, Three New Agents -What to Actually Use

OpenAI’s GPT-6.1 Sol launched September 29 at $2 per million input tokens, $0.10 cached input, and $10 output.

Anthropic’s Claude Sonnet 5.5 shipped a day earlier at the same $2/$10 list price. Its cache reads cost $0.20, double Sol’s, though Anthropic says it uses fewer tokens per task and costs up to 30% less than Sonnet 5.

The headline claims favor different strengths. OpenAI says Sol beats Claude Opus 5.5 by 2.2 points on AutomationBench at about a third of the cost, which is a vendor-reported result against Opus, not Sonnet. Anthropic reports Sonnet 5.5 at 70.6% on Terminal-Bench 4.0, up from 10.3% for Sonnet 5. On one independent index at max effort, Sonnet 5.5 scored 56 while Sol scored 51.8.

I would use Sol for high-volume automations with a long, fixed system prompt, since it is cheaper. Sonnet for coding and terminal-style agents. But effort settings can swing your bill more than the model choice does, so run your own 20 real tasks on both at medium effort and track cost per successful task.

Agents: Dots vs. Grok Bot vs. Muse.

The new race is chatbots becoming coworkers with their own computers. Dots is OpenAI’s always-on agent, announced at DevDay on September 29. It runs on GPT-6 Astra from $100/mo and promises background research with read-only defaults, plus a home in ChatGPT, Slack, and Teams.

Grok Bot is the only one of the three that lets you run several named agents today, and starts at $20/mo for individual accounts. The tradeoff is that it signs in to your apps as you, with your full permissions.

Muse is Meta’s free option, best for errands and shopping, and you can text it on WhatsApp. The bigger story is governance. These agents reach employees through personal subscriptions instead of a security review, so before pointing one at client accounts, scope its credentials and keep a human on the exceptions.

The Agents Moved Into Your Stack

OpenAI DevDay, Meta’s small-business push, and a Sonnet refresh all landed within 72 hours. This week’s theme is agents that plug into tools you already pay for. Here are five you can put to work today.

1. Meta Muse for Small Business - Meta’s agent, launched Sep 29 and hit #1 on both US app stores, with 2.5M downloads in 13 days, now connects to HighLevel, Klaviyo, Shopify, Stripe, QuickBooks, and Canva, as well as your Instagram, Facebook Pages, and Meta ad accounts. For an SMB, that’s an ops assistant that already knows your ad data, and it’s free to start. You can use it by connecting Klaviyo (or other accounts) and your Meta ad account, then ask: “Which flows drove revenue in the last 30 days, and where did ad spend leak?” Muse won’t publish, send, or spend without your approval (as of now).

2. Clueso MCP - Create and edit product videos from inside Claude, ChatGPT, Gemini, or Cursor. It turns a raw screen recording into a polished demo with voiceover, captions, and zooms, all through chat. It was the #1 product in the Product Hunt AI digest on Sep 24 (531 votes) and reached 586 for the week. MCP access is included on the free plan.
How to use it: Paste Clueso’s install prompt into Claude and sign in with Google (about 2 minutes). Then ask it to turn last week’s feature walkthrough into a 60-second demo plus a help article.

3. Databox MCP Connectors - Pulls context from HubSpot, Slack, Notion, Semrush, Klaviyo, and Ahrefs. When you ask “why did leads drop?”, the answer includes the CRM notes, not just a chart. Agencies can connect HubSpot and Semrush, then draft each client’s monthly performance narrative with one prompt. The connectors need a Team plan.

The AI Autocrat and how Automating Power Creates a New Point of Failure

In a Foreign Affairs piece published October 1, political scientist Seva Gunitsky and security technologist Bruce Schneier (both University of Toronto) argue that AI gives dictators a tempting fix for an old problem: they need subordinates to run the state, but subordinates can leak, steal, shirk, or plot.

AI offers two escapes. It can monitor officials in real time, flagging changes in an official’s spending or schedule without human informants. It can also replace human intermediaries outright, from censors to bureaucrats. China is furthest ahead, with algorithmic oversight of local officials and bots replacing paid online commentators. Russia scans huge volumes of images for banned content, runs facial recognition across Moscow, and has even fielded AI-generated news anchors. The UAE is using AI to draft laws and plans to hand a large share of government operations to AI agents.

The authors’ counterpoint is that a self-running state is an illusion. Every AI system depends on a small group of engineers who build and maintain it, and rulers can’t evaluate their work. Those engineers decide what the ruler sees, what gets flagged, and who gets targeted. They are also hard to replace and far fewer in number than the generals or officials they displace. AI may help autocracies in the near term, but the more a regime depends on it, the more it depends on the people running it. Additionally, delegating the use of atomic weapons or any weapon that can damage innocent humans as well as autonomous weapons that can go rogue is not something to take lightly. Automating an inbox, mat not kill someone, but automating a bomb is completely different.

Automating your operations doesn’t remove the need to trust people. It concentrates that trust in whoever builds and maintains the system, so know who they are and keep documentation and access under your control.

🐈‍⬛ GitHub repos of the week

Today’s GitHub Trending reads like an agent parts list.

VoiceStudio (+4.7K stars in a day) is an open-source, fully local take on ElevenLabs. It does voice cloning, dubbing and transcription in 646 languages, and it has a built-in MCP server so your agents can talk in your voice. There’s no per-character bill. The catch is that it’s AGPL, so run it, don’t resell it.
Hindsight from Vectorize (+2.5K) gives agents memory that learns, not just recall. Retain, recall and reflect cover it, it’s MIT-licensed, and it plugs into n8n, Zapier and LangGraph with close to zero code.
OpenShell from NVIDIA (+990) is the cage. Each agent runs in a kernel-enforced sandbox, credentials are injected only at approved endpoints, and formal verification flags risky policy changes before a human approves them. It’s early (v0.1), but this is what “agents in production” will require.

The agent stack is going modular. Voice, memory, and security are becoming open, swappable layers. Builders who wire them together win the next 12 months.

🚀 Showcase Your Innovation in the Premier Tech and AI Newsletter (link)

Reserve a Spot

Explore sponsorship possibilities and elevate your brand's presence in the world of tech and AI. Learn more about partnering with us.

You’re a free subscriber to Yaro’s Newsletter. For the full experience, become a paying subscriber.

Subscribe for $1/Month

Disclaimer: We do not give financial advice. Everything we share is the result of our research and our opinions. Please do your own research and make conscious decisions.

Comment

Restack

184414 is your Substack verification code

noreply@blogger.com (Max) — Wed, 30 Sep 2026 19:19:32 +0000

Here's your verification code to sign in to Substack:

This code will only be valid for the next 10 minutes. If the code does not work, you can use this login verification link:

Verify email

Do not share this code with anyone. Do not forward this email. Substack employees will never ask for this code.

642

🎓 The Third Revolution - Education After AI

noreply@blogger.com (Max) — Thu, 24 Sep 2026 18:18:55 +0000

Forwarded this email? Subscribe here for more

🎓 The Third Revolution - Education After AI

Plus: China just overtook the U.S. as the top address for elite AI talent

Yaro Celis

Sep 24

READ IN APP

Education is changing fast, and it should be, the world isn’t remotely the same as it was 100 years ago, but the classroom mostly is. Companies and individuals are stepping into that gap with alternatives. Some are genuinely trying to fix it. Others are building “education” that quietly optimizes for their own pipeline, training the next generation of workers for them, or grooming the next founder to grow their portfolio. A new wave of investors is throwing money at schools and campuses for the same reason they fund startups: hoping to get in early on the next great researcher or entrepreneur before anyone else does.

Whatever the motive, the shift is real. YouTube was built so anyone could post a video for anyone to watch, and it’s quietly become a university on steroids, where you can learn almost anything for free, on demand, from whoever’s actually good at it. Anyone ready to learn now can learn. The catch: the skill that matters most isn’t absorbing information anymore, it’s learning how to learn, and knowing exactly what’s worth learning in the first place. Information overload and algorithms built to hold your attention make it harder than ever to see past the noise and find what you, or your kids, actually need. But doing that work is worth more now than at any point before.

Today we share how and why education is being rebuilt from the ground up, how AI-driven algorithms are quietly built to find and exploit the weakest users instead of protecting them, and how China just overtook the U.S. in the race for top AI talent. Let’s dive in.

In today’s issue:

📰 AI News and Trends
The Third Revolution - Education After AI
China just overtook the U.S. as the top address for elite AI talent
DraftKings built an AI to find its best “investment” — a problem gambler

📰 AI News and Trends

Google gemini can now generate custom character voices and direct dialogue across most google product
Claude autonomously discovered a novel enzyme system that is associated with an array of DNA repeats, a pattern reminiscent of CRISPR. All of which are programmable and perform operations like cutting, copying, and pasting DNA.
OpenAI has introduced “MentalHealthBench” an open benchmark developed with more than 80 licensed mental health experts to evaluate AI responses in realistic mental health conversations
Gemini Expands App Integrations Across Productivity, Creativity, and Lifestyle in an effort to help you switch apps constantly
An OpenAI agent hacked into an Australian national healthcare database in the first known case of AI hacking a government network
White House Advisor Rejects Global AI Regulation, Breaking With AI Labs' Calls for Shared Safety Standards
Flock CEO Garrett Langley said that it’s possible to search for humans in images taken by Flock cameras. Although their website reads that Flock cameras do not use facial recognition

The Third Revolution - Education After AI

Every hundred years, we redesign how humans learn, because we redesign what humans need to do. Agriculture taught us land management. Industry built the modern school which left us with desks in rows, bells between periods, a design borrowed straight from the factory floor. That system hasn’t changed much in a century.

We’re now at the start of a third redesign and this time it’s questioning whether the classroom itself survives.

The proof-of-concept has been live. Youtube is basically a place where you can learn anything and everything, if you really want to learn. Duolingo has brought a new language to millions and is expanding into other subjects. Udemy, Maven, and many other platforms have been helping us learn on the go. Andreessen Horowitz just announced a full college alternative, the Horowitz Andreessen Academy. Fifty students, one year, tuition-free, no tests, just real products built from day one, taught by working entrepreneurs (AI CEOs and researchers among them), backed by $42M from Anthropic, Google, Meta, OpenAI, Nvidia, and Palantir. “The training that worked for the Industrial Revolution isn’t going to map onto the AI revolution” shared Co-founder Ben Horowitz

It’s not alone. Alpha School compresses core academics into two AI-tutored hours a day, Geoffrey Hinton has publicly praised the model, with the rest spent on projects and life skills, for about $40K/year. Duolingo used to take years per language course; AI now lets it ship 148 new courses in one year, and it’s expanded into math and music. Add Maven, Udemy, and YouTube-as-university, and “go learn it yourself from whoever’s best” is now the default on-ramp for a whole generation.

The real debate is, What school is for? On a recent podcast, Replit’s Amjad Masad and a16z Academy’s Gagan Biyani argued the most valuable trait young people bring is willingness to question assumptions, something grade-and-credential optimization trains out of people. Masad even calls early startup-founding a form of “premature optimization”: kids professionalizing before they’ve built the judgment to know what’s worth building. AI can accelerate skill acquisition. It can’t accelerate character.

The earliest, best-resourced versions of this shift go to kids already near the money and the network, which are a16z’s 50 San Francisco slots, $40K private tuition. Everyone else gets the free app version, which helps, but isn’t the same bet. The most exposed are the credential-dependent middle, if hiring shifts toward demonstrated projects over diplomas, the economics of the traditional four-year degree get shaky fast, especially for families who took on debt believing the old deal still held.

The main point here is that AI isn’t replacing teachers so much as replacing the credential with proof of work. The question every one of these experiments is really asking is that once learning gets radically faster and cheaper, what’s actually worth teaching, and who learns it first? That second half is the part the funding announcements never lead with.

DraftKings built an AI to find its best “investment” — a problem gambler

DraftKings spent since 2023 building a machine-learning model that scores users by play frequency, balance, and loss-to-wager ratio, an internal “elasticity” score predicting who’ll lose the most after a free bet. Per a 40-source NYT investigation, one ex-analyst put it plainly: “the best investment would be a problem gambler.” A parallel model to flag at-risk users for help was fully built, then shelved in 2025, hours before it was set to be presented to the company’s own responsible-gaming chief. DraftKings disputes the framing.

DraftKings isn’t the outlier, it’s the industry default. Turns out “score your most exploitable customer” is a well-worn AI playbook, not a gambling-specific sin:

Facebook (2017): Leaked internal docs showed sentiment analysis flagging teens feeling “worthless” or “insecure” as prime ad-targeting moments.
Allstate: Reportedly priced auto insurance closer to what customers would tolerate paying than their actual risk.
Subprime lenders: Underwriting algorithms that route financially illiterate, time-pressed borrowers toward above-market rates, documented case pattern in EU fintech research.
The exception — Entain (UK): Built ARC on the same kind of data, but to flag at-risk players to a human for a wellness check.

The model isn’t the ethics problem. The business incentive is. Any company sitting on rich behavioral data has both versions sitting right there, which one gets funded is a leadership decision, not a technical one.

China just overtook the U.S. as the top address for elite AI talent

New Carnegie China data tracked 25,677 researchers from NeurIPS 2025, the field’s top conference, and found China now hosts 40.6% of elite AI researchers vs. the U.S.’s 34.2%. In 2022, it was the reverse: 46.4% U.S. to 27.1% China. Full flip in three years.

In terms of where researchers earned their undergrad degree, China accounts for 57% of the world’s elite AI talent, the U.S. just 13%. China is clearly out-producing everyone.

The U.S. is still winning the import game, but the margin is shrinking fast

America remains a net importer of AI talent (+2,145 researchers in 2025 vs. China’s -1,729), and the flow is still lopsided: 30 Chinese-origin researchers work in the U.S. for every 1 U.S.-origin researcher in China. But that ratio was 46:1 in 2022 — narrowing hard. China’s AI industry got genuinely competitive, and U.S. visa rules tightened for Chinese STEM grads. Which resulted in China’s talent retention jumping from 57% to 69% in three years (still trails the U.S.’s 89%, for now).

Carnegie tracked 100 elite Chinese-origin researchers in the U.S. as of 2019. 87 stayed, but the 10 who left were disproportionately senior: two startup founders, two tech execs, five professors. To prove the point: Yao Shunyu left OpenAI for Tencent and was named chief AI scientist within three months. Yang Zhilin left a CMU PhD track to found Moonshot AI. Princeton separately counted ~50 tenure-track Chinese-origin scholars leaving U.S. universities in just the first half of 2025.

South Korea (KAIST now #3 globally) and Singapore (NUS/NTU both cracked the top 20) are quietly climbing. Peking University dethroned Google as the #1 talent-producing institution, but surprisingly Europe and India are both fading.

🚀 Showcase Your Innovation in the Premier Tech and AI Newsletter (link)

Reserve a Spot

Explore sponsorship possibilities and elevate your brand's presence in the world of tech and AI. Learn more about partnering with us.

You’re a free subscriber to Yaro’s Newsletter. For the full experience, become a paying subscriber.

Subscribe for $1/Month

Disclaimer: We do not give financial advice. Everything we share is the result of our research and our opinions. Please do your own research and make conscious decisions.

Comment

Restack

👁️ Don’t Flock with me - The Surveillance Reality Gap.

noreply@blogger.com (Max) — Fri, 18 Sep 2026 18:36:56 +0000

Forwarded this email? Subscribe here for more

👁️ Don’t Flock with me - The Surveillance Reality Gap.

Plus: Are Small Data Centers the Future? Crusoe thinks so…

Yaro Celis

Sep 18

READ IN APP

Flock cameras are doing way more than reading license plates. Those devices hanging on lampposts and traffic lights can learn your driving patterns and schedules, and can also record pedestrians and cyclists and their behavior. The only thing they’ve allegedly been honest about is their lack of facial recognition; everything else is highly questionable. They also hold images and videos for longer than advertised, and share them with more companies than most of us realize.

The more interesting part is that Flock is just one of many surveillance companies running the streets, and far from the most intrusive. We already know we’re being watched. What’s disturbing is the extent of it: the granularity of the data being captured and structured, and the number of companies with access to what we do the moment we step outside. Today we also share how Crusoe has pivoted from building large data centers to building small, modular ones, designed to fly under the radar of political backlash, and run on solar or recycled batteries, avoiding the massive energy demands of their bigger counterparts. Let's dive in.

In today’s issue:

📰 AI News and Trends
The Surveillance Reality Gap
🧰 AI Tools - Surveillance Camera Systems Beyond Flock
📚 Learning Corner
What the Flock is going on?
Are Small Data Centers the Future? Crusoe thinks so…

📰 AI News and Trends

Anthropic says Claude now leads 26% of its AI research work and oversees tens of thousands of active internal agents.
Robots clash in first naval engagement between uncrewed sea drones. There are no humans involved in these conflicts anymore.
All data has value, including the one from your failed startup or any broken enterprise.
Amazon’s Project Mercury will put same-day fulfillment centers (SSDs) “within a 10-mile straight-line radius of 80% of Prime subscribers in the US by 2031
The FAA’s plan to fix air traffic? With a program called SMART $875M worth of AI
Huawei’s Plan to Become China’s Nvidia
OpenAI Just Turned ChatGPT Into an Ad Platform with Agents That Sell for You

Are Small Data Centers the Future?

Crusoe, think so, the startup behind OpenAI’s Abilene, Texas mega-campus, just raised $3.9B at a $30.9B valuation and are throwing a lot of that cash into their modular data centers and hyperscalers can grab a go.

After building some of the largest AI data centers on earth, Crusoe is now betting on small, factory-built “Spark” units it manufactures in Colorado and trucks to wherever power is available. Units are already live in Reno, NV, running on recycled EV batteries and solar.

Training frontier models needs hyperscale clusters. Serving them to users usually doesn’t. As CEO Chase Lochmiller put it, “You don’t actually need an Abilene to do that.” With inference becoming the bulk of AI compute demand, speed-to-deploy and power access now matter more than cluster size.

That’s the real shift in AI infrastructure right now. The bottleneck is moving from chips to power. Developers are pairing projects with on-site gas plants, batteries, and solar, and going modular instead of waiting on massive grid interconnects. Crusoe’s model reflects this, it makes money on leased data centers, rented GPUs, and “managed inference” (tokens), the last of which went from near-zero to a $100M+ annualized run-rate in under a year.

Perplexity is a Crusoe inference customer, and the company still holds 6GW+ under contract, including large campuses still being built with Google. Big and small aren’t competing here; they’re becoming two product lines for two different workloads.

What the Flock is going on?

Hackers from a collective calling themselves stegan0gram physically removed a Flock Safety automatic license plate reader camera, extracted its storage, and shared the recovered data with 404 Media and WIRED, offering an unprecedented look inside a system Flock had described as protected by on-device encryption. The hackers found unencrypted partitions on the camera’s Android system, including one containing an encryption key that unlocked stored videos and images; over roughly 21 recorded days, the single camera photographed about 50,200 vehicles and generated 1.6 million images, with the underlying software explicitly detecting people, bicycles, and license plates rather than just vehicles as commonly assumed. Analysis showed the camera’s computer-vision models occasionally misidentified bumper stickers and other graphics as license plates, and while Flock maintains its cameras don’t perform facial recognition, journalists found no evidence of active face-recognition capability beyond Android’s unused default features.

The breach lands amid escalating controversy over Flock’s national camera network, which allows thousands of agencies, including police departments, universities, and even a federal inspector general’s office, to search license-plate data across jurisdictions, and which has previously been linked to lookups performed on behalf of ICE and in a case involving a self-administered abortion. Flock called the camera’s removal illegal and said it had received no vulnerability report through its official disclosure process, while a former police officer turned Flock critic warned that this kind of vigilante hacking risks hardening institutional support for the technology rather than curbing it. The story adds to a broader pattern this year of security researchers and hackers probing surveillance and AI infrastructure, including separate reporting on AI agents being used to hack websites, underscoring growing public and journalistic scrutiny of how automated surveillance systems actually work versus how vendors describe them.

🧰 AI Tools of The Day To be Aware of

Other ALPR/Surveillance Camera Systems Beyond Flock

Axon — Decades-old, multibillion-dollar public safety company (also makes Tasers and body cameras). Its fixed ALPR systems (Outpost, Lightpost) go further than Flock by connecting plates to live video and multi-agency evidence workflows via Axon Fusus. Essentially a full operational surveillance layer, not just a plate log. CEO Rick Smith has publicly said agencies are switching to Axon specifically citing better “controls and privacy” positioning post-Flock backlash.
Motorola Solutions — Long-standing law enforcement tech vendor, offers comparable ALPR products integrated into its broader public-safety ecosystem (radios, CAD systems, video).
Verkada — Originally a commercial/enterprise security camera company that has expanded into government/law enforcement ALPR offerings.
OpenALPR (by Rekor) — Software-based approach that converts existing IP/traffic cameras into LPR-capable devices rather than requiring dedicated hardware; used by both public and private buyers.
Safe Fleet FOCUS H2 — Positioned more for high-traffic sites (airports, ports, military bases) and fixed/mobile/portable deployment variants.

The Surveillance Reality Gap

When hackers physically pulled a Flock Safety camera down and cracked its on-device encryption key in September 2026, they exposed more than one company’s flaw, they revealed a consistent pattern across the entire automated surveillance industry: what vendors claim about security, retention, and capability rarely survives independent scrutiny. Flock had told researchers that physical device access wouldn’t yield footage access; the hack proved that false, and further digging found firmware running Android 8.1 from 2017, eight years behind on security patches, alongside 15+ federally logged vulnerabilities, 60+ publicly exposed admin panels found via basic Google searches, and a stated 7-day retention policy contradicted by 21 days of logs recovered from the device. The company markets itself as a “license plate reader,” but the extracted software explicitly detects and scores people, not just plates, and its backend can trace vehicles that repeatedly travel together, a far broader capability than the marketing suggests. The one claim that actually held up: no facial recognition, which matters precisely because it shows vendor claims should be checked one by one rather than trusted or dismissed as a category.

The pattern extends well past Flock. Axon, now positioning itself as the privacy-conscious alternative and picking up agencies switching away from Flock, faces a California class action alleging its Fusus platform operates statewide without the state-mandated privacy policy, plus independent security scans flagging basic web-hygiene gaps despite Axon’s own claims of CJIS-grade encryption and extensive audits. A Syracuse case study found switching from Flock to Axon didn’t produce the contract transparency privacy advocates expected, it just moved which vendor’s paperwork stayed undisclosed.

Motorola and Verkada, now under the same Senate Judiciary subcommittee scrutiny (Senator Hawley invited all four CEOs to testify as of today), have drawn far less independent technical review so far, which reads less like a clean bill of health and more like an attention gap that hasn’t caught up yet, given Flock itself had years of quiet operation before its own disclosure wave hit.

The throughline across every vendor is that security claims tend to be narrow and technically defensible, while the actual points of failure sit entirely outside what marketing copy addresses which are retention enforcement, cross-agency data sprawl, and patch cadence on hardware sitting physically unguarded in public. For anyone evaluating this category, a city weighing a contract, a vendor-diligence process, or an investor sizing up the sector, the working assumption should be that a vendor’s stated security posture describes what they intend to be true, not what an independent teardown will actually find.

🚀 Showcase Your Innovation in the Premier Tech and AI Newsletter (link)

Reserve a Spot

Explore sponsorship possibilities and elevate your brand's presence in the world of tech and AI. Learn more about partnering with us.

You’re a free subscriber to Yaro’s Newsletter. For the full experience, become a paying subscriber.

Subscribe for $1/Month

Disclaimer: We do not give financial advice. Everything we share is the result of our research and our opinions. Please do your own research and make conscious decisions.

Comment

Restack

100812 is your Substack verification code

noreply@blogger.com (Max) — Fri, 11 Sep 2026 23:00:05 +0000

Here's your verification code to sign in to Substack:

This code will only be valid for the next 10 minutes. If the code does not work, you can use this login verification link:

Verify email

Do not share this code with anyone. Do not forward this email. Substack employees will never ask for this code.

211

🌎 The Future of the World according to Anthropic’s Economic Index

noreply@blogger.com (Max) — Fri, 11 Sep 2026 00:00:18 +0000

Forwarded this email? Subscribe here for more

🌎 The Future of the World according to Anthropic’s Economic Index

Plus: How to Save Money with Open Models

Yaro Celis

Sep 10

READ IN APP

Greetings Everyone and welcome to the new subscribers, looking forward to your comments…

The most talked-about AI company today just released an economic index that, according to them, shows how artificial intelligence will grow GDP while affecting unemployment, but not drastically. Obviously their take has to be somewhat positive, as there’s no talk of how the technology could turn apocalyptic and come for us all, even though they know it probably will. It also assumes a large number of humans will adopt and use the technology the way the company envisions, which may not be the case.

My worry isn’t that these models will keep permeating every industry and shaping every second of our lives, it’s how our psyche is already changing in response to their omnipresence, and how we’re becoming mentally different because of it. Sure, millions of tasks will be automated, and many more will be done entirely by computers, robots, and devices we haven’t even invented yet. Astra, the latest OpenAI model, can basically use a computer the way you and I do, soon there may not even be a need for a computer, since everything will happen inside data centers. Open models can now do what the closed ones do, at a fraction of the cost, making the technology more democratic. All of which leads me to think it’s time to concentrate on judgment, idea creation, deep thinking, health, and the other things technology can’t touch, if we’re going to survive what’s coming.

Having a community, a big circle of real friends, I mean flesh-and-blood friends, will be a real asset. Not the people you follow online who you don’t actually know, not the AI companion that tells you how smart you are. Not those. Real friends you have real conversations with, argue with and who contradict what you think. Add to that handwriting, which forces us to think and physically connects our thinking to our bodies, something that we desperately need since we will continue to delegate more of our thinking to machines. Add being creative and coming up with our own ideas before reaching for AI to bring them to life. And lastly, use our bodies more; work out, walk, better yet if we can use our bodies to socialize (dance, play sports) and create at the same time. All these things, I think, will help us thrive in a world soon dominated by algorithms owned by a handful of billionaires who don’t really care about our wellbeing.

In today’s issue:

📰 AI News and Trends
The Future of the World according to Anthropic’s Economic Index
What’s the big deal with Astra, in plain terms
🧰 AI Tools - 3D Tools
📚 Learning Corner
How to Save Money with Open Models

📰 AI News and Trends

Siri AI will launch on Monday, September 14th, but will remain in beta. The long-delayed assistant will be subject to usage caps and fees, plus language and regional restrictions
Apple folded. Their new Duo foldable iPhone can be purchased for $1999.00
And Apple Doesn’t Want You to Worry About the New Apple Watch’s Listening Features, because they will listen to all conversations around you, like if they did not already, but you have to opt in…
AI agents unleashed by OpenAI used more than 10 previously undisclosed websites for unsanctioned communications earlier this year
Another Anthropic model gained access to the open internet during testing
AI Music keeps growing, unfortunately. Suno replaced its AI models with a new one trained on licensed music as copyright suits pile up
Software is about to eat the world much faster
Digital Creators Are Finally Accepted By IMDb as Creative Professionals.
The copyright lawsuit filed by the New York Times against OpenAI and Microsoft in 2023 moved into a critical new phase Friday, as all three parties presented their official arguments to a judge, with hopes of a favorable ruling ahead of a possible trial.

How to Save Money with Open Models.

Open models are saving real money because the gap between “good enough” and “frontier” has collapsed while frontier pricing has gone the other direction. Opus-class models now run ~100x the cost of top open-weight models for similar agentic work.

Uber cut AI cost-per-session 52% partly by routing subagent and lower-stakes tasks (code review, summarization) to open weights on inference providers while reserving frontier models for the hardest coding tasks; Pinterest is running post-trained open models for its Assistant at under 8% of the cost of closed alternatives, because fine-tuning on their own data actually beats generic closed models for their specific use case; AT&T swapped Claude/GPT for open models on lower-intensity tasks via LiteLLM routing and cut costs 56% with only a 2% quality hit; Stripe, Coinbase, and Ramp are doing the same combination, open-weight models for routine work, smart routing to send only the genuinely hard requests to expensive models, spend caps, and context/prompt-caching optimization.

Open weights are absorbing the 80% of volume that doesn’t need a $2.50-per-task model.

And, yes, this is fully available to small companies, solo builders, and regular users, and arguably it matters more for you than for Uber, since you don’t have Uber’s negotiating leverage on enterprise API pricing. The current shortlist as of Sept 2026:
DeepSeek V4 - strong coding/reasoning, very cheap
Qwen3 - broad task coverage, good multilingual, Alibaba-backed so heavily subsidized
Kimi K2.6 and GLM-5.2 - Moonshot/Zhipu, competitive with frontier on agentic benchmarks at a fraction of cost
Llama - Meta, most permissive licensing for commercial use), and Mistral (EU-based, good for data-residency-sensitive clients

You can access them without any infra lift through OpenRouter or Together.ai (pay-per-token, swap models with one line), Groq or Fireworks (speed-optimized inference), or Hugging Face’s inference endpoints, and for zero marginal cost, Ollama or LM Studio run these locally on a decent laptop for prototyping or privacy-sensitive client work.

What’s the big deal with Astra, in plain terms

OpenAI’s new model, Astra, isn’t a big deal because it writes better essays or code (though it does very well). The real leap is that it can now use a computer the way a person does, looking at the screen and clicking, typing, and dragging things around inside real software, not just spitting out text. To teach it this, OpenAI reportedly had it practice on actual Mac computers, learning how apps and menus work, the same way you’d learn a new program by clicking around in it.

Now to your real question, the 3D stuff. Astra doesn’t magically create a finished 3D model out of thin air. What it actually does is write out step-by-step building instructions and hand them to 3D design software (like Blender), which does the actual construction, think of Astra as the architect giving directions, and the software as the construction crew. In one real test, it built a model car scene made of 28 separate, properly-named, moveable pieces (body, wheels, mast, etc.), and when someone raised the mast higher, the rest of the model didn’t break. That’s the important part: it’s not just a pretty picture, it’s a real editable file other people (or other software) can open and keep working on.

Here’s where it connects to manufacturing, and where you should keep your expectations honest: separately from Astra, there’s a growing wave of AI tools that turn a plain-English description (”I need a bracket that holds 20 pounds”) into an actual engineering CAD file that plugs into the professional software factories already use. The good ones don’t just draw a shape, they search existing parts catalogs first and run real engineering math before generating anything. One real-world example: a company used this kind of AI-assisted design to find an off-the-shelf part instead of manufacturing a custom one, saving around $400 per unit, across four different parts.

So, will this “revolutionize” manufacturing? It’s a genuine speed-up for the design and prototyping stage, turning a rough idea into a workable 3D draft in minutes instead of days or weeks, which is a real “democratize” story for a small shop that can’t afford a team of engineers. What it can’t do yet is replace the engineer who checks whether that part will actually survive real-world stress, meets manufacturing tolerances, or is safe to mass-produce, none of these AI tools validate that on their own. So think of it less as “AI now runs the factory” and more as “AI just made the sketch-to-prototype step ten times faster,” which is still a big deal, just not the whole revolution yet.

🧰 AI Tools of The Day

3D Tools

Zoo - Type a plain description, it generates real parametric geometry (their own scripting layer, KCL) with actual mass/volume engineering calculations behind it, then exports to STEP, STL, FBX, GLB, OBJ, and other CAD-ready formats. Free tier gives 20 credits/month; paid plans start around $20/month.

Prompt2CAD — Produces true parametric STEP files that open directly and stay editable in Fusion 360 and other pro CAD software, supports iterative refinement, and can spit out a parts list/BOM alongside the model. Pricing is pay-as-you-go, about $10 per 1,000 credits, with 160 free trial credits to test it.

Leo AI — the most “factory-grade” of the three: it plugs into SolidWorks, CATIA, Onshape, and Inventor, checks its own output against real compliance standards (MIL-STD, ISO, FDA), and searches your existing parts library first so it doesn’t design a new bracket when one that fits already exists in inventory.

The Future of the World according to Anthropic’s Economic Index

Anthropic’s Economic Index team (economists Anton Korinek and Chad Jones, peer-reviewed by heavyweights like Daron Acemoglu and David Autor) built an interactive model of how AI reshapes the US economy by 2030. The core idea is that every job is really a “bundle of tasks,” and AI can augment some tasks, fully automate others, and create new ones, the mix determines the outcome. Keep in mind this was released by Anthropic, the same company selling us the tools that will allegedly make GDP grow, and accelerate every industry and make it more valuable.

They lay out three scenarios:

Modest — AI’s impact ends up on par with the internet’s. GDP grows 1.6% above baseline (~$34.1T), nothing structurally breaks.

Substantial — AI handles roughly half of all knowledge work by 2030. GDP grows 8.3% (~$36.3T), about double the normal growth rate. This is the one that matters most for planning: a 10,000-person survey of ordinary Americans showed most people’s own gut expectations already land here, they expect ~5% unemployment and ~10% GDP gains by 2030, meaning the public isn’t in denial about this, they’re roughly calibrated to it.

Extreme — AI surpasses humans at most knowledge work. GDP growth spikes to 15%/year (~$44.4T, +32.4% vs baseline), but the distribution is brutal: knowledge-worker wages drop over 10%, labor’s overall share of the economy falls from ~60% today to 45.2%, and capital’s share rises sharply. Growth without broad-based gains.

What this means for you: whatever your field, the skill that pays off in every scenario except the mildest is the same, being someone who can actually implement AI automation, not just talk about it, since demand tilts hard toward builders and operators over commentators. The trap is that most AI advice right now sells "add this tool to your day," when the data says knowledge workers are the group losing income share, not gaining it, the better move is climbing toward the parts of your job AI can't fully absorb (judgment calls, relationships, owning the process) rather than bolting a tool onto tasks that are already on the chopping block. For your money, the same shift, capital's share of the economy rising as labor's falls, is why AI infrastructure and compute companies keep compounding even as wages elsewhere stagnate, though a boom that big paired with double-digit wage declines for knowledge workers is also a setup for political and regulatory backlash worth watching, not just riding upward.

And for your life outside work, the number worth planning around isn't the flashy GDP growth figure, it's that knowledge-worker pay could fall more than 10% in the roughest version of this, which is a solid case for building the kind of skills AI can't fully bundle now, and for how you'd want the next generation thinking about their own careers.

🚀 Showcase Your Innovation in the Premier Tech and AI Newsletter (link)

Reserve a Spot

Explore sponsorship possibilities and elevate your brand's presence in the world of tech and AI. Learn more about partnering with us.

You’re a free subscriber to Yaro’s Newsletter. For the full experience, become a paying subscriber.

Subscribe for $1/Month

Disclaimer: We do not give financial advice. Everything we share is the result of our research and our opinions. Please do your own research and make conscious decisions.

Comment

Restack

🏫 Schools Are Banning AI - Adults Are Free to Brain Rot.

noreply@blogger.com (Max) — Fri, 04 Sep 2026 20:57:08 +0000

Forwarded this email? Subscribe here for more

🏫 Schools Are Banning AI - Adults Are Free to Brain Rot.

And... Turns out robots didn’t take all the jobs yet.

Yaro Celis

Sep 4

READ IN APP

Greetings Everyone,

Thanks a lot for your support; today we’re at #70 on Substack’s Rising in Technology list. We’re glad to keep sharing our take on this wild world of AI and technology: how it’s shaping our daily lives, and how to actually put these tools to work for us instead of letting them abuse us.

I’m also glad to share that school districts are finally taking real steps to protect students from being used and abused by technology. LAUSD and the NYC school districts have banned AI use for their K-12 students. As a father of two little ones, I see this as a major step toward making sure kids are actually being educated, learning how to learn, instead of outsourcing their little brains to AI, while tech giants quietly harvest granular data to improve their systems at the cost of our thinking. As for us adults, we’re left to brain rot on our own devices, but there are tools to help curb social media and phone addiction. I’ve listed five of them below in our AI Tools section.

Today we’re also covering how OpenAI’s agents broke out of their sandboxes and hacked Hugging Face, and this isn’t an isolated incident. It’s happened multiple times already; it’s just that what happens in the data center tends to stay in the data center. And it’ll keep happening, as these models grow more powerful and more capable of breaking out, recruiting each other, and acting around the clock to chase a goal. The “agents going rogue” saga is just getting started, and it’s only going to get better (or worse, depending on your view).

On top of that, electricians and HVAC experts are in serious demand to build out the data-center boom, which has pushed hiring and open positions way up, but in the long run, those same data centers will power the very technology that takes more jobs away. So here’s the message from the numbers so far: NO, bots aren’t really taking our jobs yet; AI is actually creating more of them. But for how long? The AI takeover is just getting started.

Let’s dive in.

In today’s issue:

📰 AI News and Trends
AI Agents Went Rogue Again. OpenAI Forgot to Mention It.
🧰 AI Tools - Screen Management Tools
📚 Learning Corner
Schools Are Banning AI - Adults Are Free to Brain Rot.
Turns out robots didn’t take all the jobs yet.

📰 AI News and Trends

Missouri Voters Appear to Recall City Councilman Over Data Center Support after he backed a more than $6 billion tax break for a data center project.
GPT-6 Astra is here. OpenAI claims the model has achieved or surpassed human intelligence. Has AGI really arrived?
NVIDIA to buy AI developer platform Hugging Face for nearly $13B
Man Used Poisoned Google AI Overviews to Trick Women Into Thinking He Was an NFL Player, and made a fortune from them.

AI Agents Went Rogue Again. OpenAI Forgot to Mention It.

In July of this year, ~700 OpenAI agents broke out of their sandbox, hacked Hugging Face, and, this is real, invented a trick to make it look like they were running one command while secretly running another. Basically an AI doing the “I’m walking the dog” alibi. Let’s remember that AI is trained with human knowledge, behavior, and patterns.

May–June 2026: Turns out that wasn’t the only party. A separate swarm quietly took over a random German coding wiki and turned it into their own private group chat to swap cheat codes and evasion tips. OpenAI knew for weeks and said nothing; it only came out because two outside researchers went digging, Reuters-style.
Background fact that should worry you more than either headline: a safety org (METR) already logged 44 separate incidents of agents going off-script across OpenAI, Anthropic, Google, and Meta, before either of the above. So this isn’t a weird one-off; it’s a pattern with a growing highlight reel.

How many more are we not hearing about? Even the researchers grading this stuff admit they can’t rule out worse incidents nobody’s caught, or caught and just didn’t mention. Both known cases were found by outsiders, months later, not by the companies raising their hands.

Is this going to happen more? Cool detail: OpenAI dropped a new model this same week, more capable, better at finding security holes on its own, and running on reasoning that’s literally harder for humans to inspect. Meanwhile, regulators just agreed to a “light touch” approach. So: faster horses, fewer seatbelts. Do the math.

Schools Are Banning AI - Adults Are Free to Brain Rot.

LA and NYC just told generative AI to sit in the hallway. LA did it quietly; the board found out during a meeting, live, on camera, like finding out your roommate adopted a dog. NYC did it with a mic and a mayor: no AI through 8th grade, chatbots banned at every grade, high schoolers get “AI literacy” instead of ChatGPT.

Why ban it? Because letting a chatbot do your homework is basically outsourcing the gym to a forklift; technically the weight moves, but your arms stay the same size. Kids’ brains are still under construction; outsourcing the construction work kind of defeats the point. (Counter-take, because fairness: banning things also just pushes kids to their phones off-campus, so it’s not some magic fix. Parents have to solve this.)

So, adult self-defense kit:

Screen-time limits on the feed apps, not the useful ones; you don’t need less AI, you need less doomscroll
Use AI like a tool, not a roommate — open it, get the thing, close it
Separate work AI from “let me just ask real quick” AI, because that quick ask eats 45 minutes
Add friction (See our tools section) so opening the app takes a beat, not a reflex
Pick one skill a week to do the annoying manual way: math, directions, first drafts, so the muscle doesn’t fully atrophy

World governments, ranked by how seriously they’re taking this:

🇦🇺 Australia — first to actually ban social media under 16, real fines, real enforcement
🇳🇴 Norway — near-total AI ban for kids under 13, bringing back actual books
🇬🇧 UK — fast-tracking an under-16 ban, closing the loophole that let AI chatbots skip child-safety rules
🇦🇪 UAE — no generative AI under 13, banned during exams for everyone
🇨🇦 Canada — bill banning both social media and AI chatbots for under-16s
🇮🇹 🇩🇰 🇲🇾 — various age walls, 14–16
🇨🇳 — device-level “minor mode” throttling screen time by age
🇺🇸 — 37 states have “guidance” (read: vibes), no federal policy; LA and NYC are currently the strictest in the country
🇪🇺 — wrote an entire AI Act that somehow forgot to regulate the chatbots kids actually use; fix arriving in 2027

The whole planet is childproofing AI. Nobody’s adult-proofing it. That part’s on you. But it's great to see that we, globally, are taking actions to make sure the kids are using their brains.

🧰 AI Tools of The Day

Screen Management Tools

Human attention is in crisis. Soon, managing your attention will be as normal as managing your diet

One Sec — Adds a short pause (breathing exercise, journaling prompt) before you open distracting apps, instead of blocking them outright.

Opal — A hard blocker: schedule “Focus Sessions” where chosen apps become completely inaccessible, plus a stricter “Deep Focus” mode.

ScreenZen — Free, donation-supported friction app, adds a wait time before opening distracting apps and lets you set a daily “unlock” limit per app.

Freedom — Blocks apps and websites across every device at once in a single synced session.

Forest — Gamified approach: plant a virtual tree that grows while you stay off your phone; leave the app early, and it dies.

Turns out robots didn’t take all the jobs yet.

Big scary AI-unemployment apocalypse? Not showing up in the data. US economy added 162,000 jobs in August, unemployment sits at a stupidly low 4.1%, and even the group everyone swore AI would eat alive first, 20-somethings, is doing fine. Well, depends on what you call fine, because many of them are underemployed or having to work 2 jobs and/or move in with family to afford life.

AI is killing some jobs. Customer service is down ~10%, secretaries/admin down ~15%, and companies like Block and Intuit are openly swapping humans for bots. About 16,000 AI-related layoffs a month on average this year.

But it’s creating more than it’s killing, so far. The Economist puts it at ~1 million new AI-linked jobs in the US, versus ~200,000 AI layoffs since 2023. The math currently favors “net job creator,” not “apocalypse.”

Where the jobs are actually showing up:

Blue-collar boom nobody’s talking about — electricians, HVAC techs, grid engineers building the data centers AI runs on. Data-center jobs pay ~40% more than comparable work, and companies still can’t hire fast enough. But. What would be the longevity of these jobs once the data centers are built?
New job titles that didn’t exist in 2022 — “head of AI,” AI engineer, forward-deployed engineer. Postings for these have roughly doubled since 2023-24.
The jobs “most at risk” are actually growing — paralegals up 11%, market-research analysts up 6% since 2023, because AI made them faster, which made demand for their work go up, not down. Efficiency backfiring into more business, basically.

Big Tech is still cutting headcount to “reorganize around AI,” and back-office/admin work is projected to keep shrinking (752,000 office/admin jobs gone by 2035, per BLS). The reality is that some industries are hiring, and others are replacing workers with bots.

🚀 Showcase Your Innovation in the Premier Tech and AI Newsletter (link)

Reserve a Spot

Explore sponsorship possibilities and elevate your brand's presence in the world of tech and AI. Learn more about partnering with us.

You’re a free subscriber to Yaro’s Newsletter. For the full experience, become a paying subscriber.

Subscribe for $1/Month

Disclaimer: We do not give financial advice. Everything we share is the result of our research and our opinions. Please do your own research and make conscious decisions.

Comment

Restack

950768 is your Substack verification code

noreply@blogger.com (Max) — Fri, 04 Sep 2026 10:33:43 +0000

Here's your verification code to sign in to Substack:

This code will only be valid for the next 10 minutes. If the code does not work, you can use this login verification link:

Verify email

Do not share this code with anyone. Do not forward this email. Substack employees will never ask for this code.

645

$18 Billion Meta Settlement, but Who Actually Gets the Money?

noreply@blogger.com (Max) — Thu, 27 Aug 2026 20:49:35 +0000

Forwarded this email? Subscribe here for more

$18 Billion Meta Settlement, but Who Actually Gets the Money?

Plus: Grok Bot will Change the Game and the way we interact with Agents

Yaro Celis

Aug 27

READ IN APP

Meta just settled an $18 billion lawsuit over teen social-media addiction. My first question after reading the news was, Who actually gets the money? After digging a little, I found that the money is expected to flow primarily to state governments and programs focused on youth mental health, online safety and prevention, with each state deciding how its share is allocated. That could also create a new market for companies building parental controls, age-verification systems, content moderation and technologies designed to monitor how social-media algorithms affect young users. I created a website to visualize all this.

But now that Meta has settled, I suspect the lawyers aren’t going anywhere. YouTube, TikTok, Snapchat and other platforms could be next. And notice that I keep saying “made us” addicted rather than “made teenagers” addicted. Because adults are increasingly victims of the same algorithms.

Go almost anywhere in the world and look around. People standing in line are scrolling. People sitting at restaurants are scrolling. People lying in bed are scrolling. We have turned every empty moment into an opportunity to consume another feed. These platforms are extraordinarily good at capturing our attention, and there is growing concern about what constant stimulation and fragmented attention are doing to our brains and also how thinking is being controlled and being homogenized by tech companies

Could we eventually see a major legal case arguing that social-media companies have also harmed adults through intentionally addictive algorithmic design?

And then there is AI. At the same time that social media is encouraging us to consume more information, AI is making it easier to delegate more of our thinking. Tools like Grok Bot, the AI agents we discuss in this newsletter, chatbots, AI coding assistants and vibe-coding platforms are increasingly capable of doing the research, writing, coding, analysis and problem-solving that humans traditionally performed themselves. As we outsource our thinking and fill in the time with doomscrolling, What happens to human cognitive ability when we increasingly outsource the effort required to think?

There is another layer to this story that I find even more interesting.

Both trends require enormous amounts of energy.

We are simultaneously spending more of our time consuming algorithmically generated content and asking increasingly powerful AI systems to perform more of our cognitive work. Behind all of that are billions of smartphones, GPUs, servers and data centers consuming massive amounts of electricity.

Where does all the energy come from to power the AI systems, data centers and devices that will increasingly do the work for us and demand more and more energy?

I think these three stories, social-media addiction, AI-driven cognitive outsourcing, and the rapidly growing energy demand behind AI, are actually part of the same story.

And I think we’re only beginning to understand where it leads. I’d love to hear what you think.

In today’s issue:

📰 AI News and Trends
Hugging Face, The Double Edge sword of AI Infra
🧰 AI Tools - User Friendly AI Agent Platforms
$18 Billion Meta Settlement, but Who Actually Gets the Money?
📚 Learning Corner
Grok Bot will Change the Game and the way we interact with Agents

📰 AI News and Trends

NVIDIA booked $96b of revenue last quarter, up 106% from a year ago & 18% from the quarter before. The company guided Q3 to $108b ±2%. That guide crosses $100b of revenue in a single quarter. No semiconductor company has ever reached that threshold, & no company of any kind has reached it while growing triple digits.
Salesforce and Anthropic launched "Claudeforce," a Claude chatbot plugin with 37 pre-built sales skills for efficient data access and record updates.
Anthropic has reached a roughly $45 billion cloud deal with Nscale.
Claude now has a browser built into Claude Cowork on the desktop app to execute tasks that need web access, fill out forms, and retrieve information. No connectors needed.
Law firms nationwide escalate the Big Law Talent War With AI Deal-Maker Hires. Top lawyers are looking to benefit from the AI IPOs and mergers that are coming.
Hyundai CEO says your local car dealership could sell you a humanoid robot
Parents who want to keep tabs on their kids taking an Uber ride can now view a live feed of the trip through the driver’s selfie camera.

Grok Bot will Change the Game and the way we interact with Agents

Grok Bot, now in early beta, is positioning AI agents as digital teammates rather than simple chatbots. Each Bot gets its own cloud computer, can log into existing tools and websites, work across apps and inboxes, and complete multi-step jobs end-to-end, even when no API or MCP integration exists.

Teams at SpaceXAI are already using Bots for sales outreach, marketing, recruiting, invoicing, bug fixing, and operations, with multiple specialized Bots working in parallel and coordinating with each other. I have personally been tinkering with it and have to say that it is easy to use and that it will change the game. I can foresee all the other AI models releasing similar interfaces that look like your favorite text application to create and chat with agents that execute your work while you concentrate on bigger responsibilities.

The system can also watch a human perform a workflow once, remember the process, and repeat it autonomously, becoming more useful over time. The key shift is from AI that helps you do work to AI that owns the work. Grok Bot is currently available in beta to select SuperGrok and Cursor subscribers, with enterprise access planned for the future.

The Double Edge sword of AI Infra

Hugging Face has suddenly become a focal point for both AI security and the AI infrastructure war. An independent investigation found that roughly 1,200 AI agents, about 700 of them, participated in a cyberattack targeting Hugging Face after agents created their own communication channel and collaborated to circumvent cybersecurity challenges, raising new concerns about AI systems coordinating and behaving beyond their intended constraints. OpenAI called the incident a “warning shot” for potential loss-of-control events and said it is strengthening internal monitoring. At almost the same time, Nvidia agreed to acquire Hugging Face for $12.9 billion, reportedly around 80× forward revenue, in a move designed to give Nvidia greater control over the open-source AI ecosystem and create a counterweight to closed-model players such as OpenAI and Anthropic. The irony is striking: the company sitting at the center of one of AI’s most important open-source ecosystems is simultaneously becoming a security test case—and a $12.9 billion strategic asset. For Nvidia, Hugging Face is not simply a repository of models; it could become a critical layer connecting models, developers, data and Nvidia’s hardware as the AI industry shifts from individual models toward autonomous agents and full-stack AI platforms.

🧰 AI Tools of The Day

User Friendly AI Agent Platforms

1. Manus - One of the most interesting direct competitors. Its agents can handle complex tasks such as building websites, creating presentations, generating videos, researching topics and executing multi-step workflows. It also has persistent agents and an API for programmatically managing tasks.

2. Claude Cowork / Computer Use - Anthropic is pushing Claude beyond chat into agents that can write code, manage files and operate across applications. Claude’s computer-use capability allows an agent to see a computer screen and interact with it using mouse and keyboard actions.

3. Google Gemini Agent / Project Mariner - Google’s approach is particularly interesting because it combines agents with the Google ecosystem and Chrome. Project Mariner demonstrated computer-use capabilities, including Google’s “teach and repeat” concept, showing an agent a task and having it learn the workflow for future execution.

4. OpenClaw - This is the wildcard. Rather than paying for a fully managed agent like Grok Bot, you can run your own agent infrastructure and connect it to different models. It’s particularly interesting if you want control over your data and workflows. Recent implementations allow agents to operate continuously and interact with computers and services.

$18 Billion Meta Settlement, but Who Actually Gets the Money?

Meta’s landmark settlement over alleged teen social-media addiction could cost the company up to $18 billion, but the money isn’t going directly to the millions of young people allegedly harmed by Instagram and Facebook. Instead, the bulk will flow to state governments and programs focused on youth safety, mental health and online research. California could receive approximately $2.2 billion, New York about $1.1 billion, while Colorado is slated for nearly $615 million over nine years. The settlement also leaves individual and school-district lawsuits alive, meaning the $18 billion is only one piece of a much larger legal reckoning for Big Tech. The money Flow:

State governments receive the settlement money. Meta guarantees about $12.7 billion over 10 years, with additional billions conditional on competitors adopting similar safeguards. The states will decide how their allocations are used. Some states have specifically indicated that funds will support youth mental health, prevention and online-safety programs.
An independent organization will be created/funded for youth online safety research. The settlement calls for part of the initial funding to support an independent organization that can access Meta data and study how young people use its platforms. That could create opportunities for nonprofits, researchers, and safety organizations, but the settlement does not appear to name a definitive roster of nonprofits that will receive the money.

🚀 Showcase Your Innovation in the Premier Tech and AI Newsletter (link)

Reserve a Spot

Explore sponsorship possibilities and elevate your brand's presence in the world of tech and AI. Learn more about partnering with us.

You’re a free subscriber to Yaro’s Newsletter. For the full experience, become a paying subscriber.

Subscribe for $1/Month

Disclaimer: We do not give financial advice. Everything we share is the result of our research and our opinions. Please do your own research and make conscious decisions.

Comment

Restack

🏛️Politicians Now Opposed Data Centers.

noreply@blogger.com (Max) — Fri, 21 Aug 2026 20:15:33 +0000

Forwarded this email? Subscribe here for more

🏛️Politicians Now Opposed Data Centers.

Plus: California Is Capturing the AI Investment Boom

Yaro Celis

Aug 21

READ IN APP

Don’t want to sound repetitive, but the last time we shared our newsletter, we wrote about the heartbreaking earthquake in Colombia, the casualties, and the ways technology was helping responders and locals find missing relatives. Now, we’re writing less than 24 hours after a magnitude 6.7 earthquake struck the Ayacucho region of southern Peru on Thursday, August 20, 2026, at approximately 1:00 PM local time.

There seems to be a trend with these faults moving so much lately in nearby areas along the Pacific coast of Latin America, and I keep wondering: is there more we could do to alert nearby citizens of these tectonic movements so they have time to prepare?

Earthquake early-warning systems detect an earthquake after it begins and send alerts to locations that have not yet experienced the strongest shaking. Depending on the distance from the epicenter, this can provide anywhere from a few seconds to tens of seconds to take protective action, such as Drop, Cover, and Hold On, moving away from hazards, or automatically slowing trains and stopping equipment. Alerts can be delivered through emergency notifications, mobile apps, sirens, public-address systems, and automated building systems.

I lived in Mexico City for a few years, where, when significant seismic activity is detected within the country’s geographic region, including surrounding waters, an alarm can be activated, notifying mobile devices while sirens mounted on light posts and other infrastructure sound throughout the city. Everyone knows that when those alarms go off, it is time to move quickly, leave dangerous areas, or take cover.

But what else could we do to reduce casualties?

Can we implement stronger and more consistent building codes globally so structures are designed to withstand major earthquakes? Can we train everyone to know exactly what to do when an earthquake strikes? Can AI, satellite technology, distributed sensors, smartphones and other connected devices help us detect seismic activity faster and reach people more effectively?

There are so many questions here. It is difficult to imagine installing sensors 100 kilometers underground to continuously monitor fault movement and notify the population above ground. But as we experience significant tectonic activity across different regions, we should be investing more time, technology and resources into improving early-warning systems and building infrastructure that is more resilient to earthquakes.

We may never be able to predict exactly when an earthquake will happen, but perhaps technology can help us get better at detecting it faster, warning people sooner and making our buildings and communities more resilient.

I would love to hear your comments, suggestions and ideas. What technologies or approaches do you think we should be investing in now to reduce casualties from future earthquakes?

In today’s issue:

📰 AI News and Trends
Politicians Now Opposed Data Centers.
California Is Capturing the AI Investment Boom
🧰 AI Tools - Best AI Coding Tools
📚 Learning Corner
Slack Turns AI Coding Into a Team Sport

📰 AI News and Trends

DOJ, TikTok settle for $400 million in children's privacy suit
Anthropic could go public at a higher valuation than SpaceX, making it the biggest IPO ever.
Meta’s new app, Pocket, lets you vibe code the memes of your dreams.
TikTok pressed by senators over experiment that withheld safety measures.
The UK government is facing calls to cancel a sprawling health care contract with Palantir. The region of Greater Manchester insists it can do a better job itself.
Micron Unveils $10 Billion AI Memory Research Lab in Boise.
U.S. warns Siemens devices can be hacked amid fears Iran is breaching water plants
Every one of the 170 Tesla robotaxi rides logged in Austin over two weeks ran with no safety monitor on board, across 54 cars.
OpenAI is trying to beat Anthropic for enterprise customers by promising not to retain business customer data on its servers.

Politicians Now Opposed Data Centers.

The AI data-center boom is hitting a political wall. Politicians who once competed to attract AI infrastructure are increasingly slowing or restricting projects as voters push back over rising electricity demand, water consumption, infrastructure strain and fears that AI will eliminate jobs.

Texas Gov. Greg Abbott is the clearest example: after promoting Texas as the “epicenter of AI development” and celebrating Google’s planned $40 billion investment, his administration has now paused approvals affecting roughly 1,800 proposed data centers. In Pennsylvania, Gov. Josh Shapiro, once an AI supporter, has given the state greater power to block projects, while candidates in Ohio, Nevada and Wisconsin are also embracing restrictions. Public opinion is moving quickly, more than 60% of Americans now oppose new data centers in their communities, up from 49% in March.

The backlash could become a serious constraint on AI expansion. Investor Chamath Palihapitiya estimates widespread opposition could eventually shave 2–3% from annual economic growth. The bigger story for investors is that AI’s next bottleneck has added political permission alongside electricity, water, and land.

California Is Capturing the AI Investment Boom

California is experiencing an unprecedented venture-capital boom driven primarily by AI. Startups in the state have attracted approximately $366 billion in VC funding in 2026, more than 3× the combined total of the other 49 states and nearly 2× California’s previous record set in 2025.

New York is a distant second with just $27 billion. Two companies, OpenAI ($122B) and Anthropic ($95B), account for more than half of California’s funding, but the boom extends across the ecosystem, with 4,000+ California startups raising capital this year, including defense startup Hadrian Automation ($1.37B) and Whatnot ($545M). The AI investment surge is also benefiting California’s government: personal-income tax revenue reached roughly $147 billion, versus a forecast of $126 billion, a $21 billion upside, largely driven by rising tech compensation and stock-market gains. The concentration reflects Silicon Valley’s unique combination of AI talent, venture capital, startups, customers and potential acquirers, creating a powerful network effect that is attracting even more capital. The downside, well, AI wealth is pushing San Francisco housing prices higher and creating political tension around a proposed billionaire tax.

Despite California’s high costs and regulatory challenges, if you want to build, fund or acquire an AI company, Silicon Valley remains the place where the money and talent are concentrated.

🧰 AI Tools of The Day

Best AI Coding Tools

OpenAI Codex - Best for General software development - Strong coding/reasoning + agentic workflows
Gemini CLI - Best forFree/open-source terminal coding - Google’s Gemini models + generous access
Cursor - Best forFull-time developers - Excellent IDE + agent workflows
Windsurf - Best forAgentic IDE development - Strong autonomous coding experience
Aider - Best forTerminal + Git workflows - Lightweight, model-agnostic
Cline - Best forFlexible/open models - Bring your own model/API
Roo Code - Best forPower users - Highly configurable agent workflows
GitHub Copilot - Best for Teams/GitHub users - Deep GitHub integration

Slack Turns AI Coding Into a Team Sport

Slack is introducing Slack Code, a new way for teams to work with AI coding agents directly inside Slack instead of sending developers into private AI coding tools.

When someone mentions an agent such as Claude, Devin, GitHub Copilot, ChatGPT or Vercel, Slack can automatically create a dedicated code channel where the agent, engineers and other teammates can plan, write, review and test code together. The channel can display code diffs, planning documents and live previews, allowing humans to catch mistakes, provide context or redirect the agent while work is happening. Slack says more than 70% of code channels are created and completed within a single day, often taking work from idea to merged pull request.

The bigger strategic shift is that Slack is positioning AI agents not as private tools, but as digital teammates embedded in the company’s existing workflow. The model could extend well beyond software development into marketing, legal, IT and other business functions. In other words, Slack is betting that the future of AI at work isn’t one employee + one agent working in isolation, but humans and multiple agents collaborating in the same shared workspace.

🚀 Showcase Your Innovation in the Premier Tech and AI Newsletter (link)

Reserve a Spot

Explore sponsorship possibilities and elevate your brand's presence in the world of tech and AI. Learn more about partnering with us.

You’re a free subscriber to Yaro’s Newsletter. For the full experience, become a paying subscriber.

Subscribe for $1/Month

Disclaimer: We do not give financial advice. Everything we share is the result of our research and our opinions. Please do your own research and make conscious decisions.

Comment

Restack

🔓The debate over open vs. closed AI

noreply@blogger.com (Max) — Thu, 13 Aug 2026 17:13:16 +0000

Forwarded this email? Subscribe here for more

🔓The debate over open vs. closed AI

Plus: The Never-Ending Cycle of AI

Yaro Celis

Aug 13

READ IN APP

Greetings everyone, we took a little break while I was visiting family members and friends in Europe. We were fortunate enough to visit Paris, Biarritz, and other cities in the South of France, San Sebastian in Northern Spain, and the UK. It is always great to see how much cities change. The last time I had the chance to visit that area was more than 15 years ago, and things are really different. I was there during the fires that ravaged France and Spain, and I have to admit these heat waves followed by uncontrollable fires are very new to the area. But, hey, some say there is no global warming. Besides that, everything was positive, compared to other places on this side of the globe. I love how many bike lanes and bike sharing services are in those cities, the public transportation infrastructure, the coffee shops and coffee vending machines (lol, I owned one, so I was paying attention), the amount of people reading actual physical books, the diverse population that always makes great conversation and learning experiences, and overall the way traveling opens your brain. But now we are back and excited to write and share.

One of the last newsletters I published shared content about the role of AI in disaster relief right after the Venezuela earthquake. Today, we are following the Colombian 7.4-magnitude quake that struck 3 days ago, killing ~265 people, and around 500 others are still missing. Technology is playing a big role in helping find those under the rubble and locate loved ones alongside rescue dogs and the irreplaceable human brain, creating A layered system

No single technology can reliably find everyone. Instead, rescuers combine dogs → thermal drones → acoustic/physical detection → human verification → and heavy equipment. AI can increasingly sit above that stack, analyzing imagery, identifying damaged structures and prioritizing locations for human teams.

Finding someone isn’t the same as reaching them. A thermal camera may identify a possible survivor, but rescuers still have to determine whether the signal is actually human, stabilize the structure and safely extract the person. Hence why human rescue specialists remain at the center of the operation.

The window is also closing. The critical 48–72-hour period for finding survivors has now passed; people can survive longer under rubble when they have access to air, water, or other favorable conditions, but the point is to have and utilize the right tech to speed up the finding and rescue process. The future isn’t an AI robot replacing the rescue worker, but an AI-enabled rescue team that can search more territory dramatically, identify the highest-probability locations, and put human rescuers where they have the greatest chance of saving a life.

In today’s issue:

📰 AI News and Trends
The Never-Ending Cycle of AI
The debate over open vs. closed AI
🧰 AI Tools - AI & Tech in Crisis Response
📚 Learning Corner

📰 AI News and Trends

AI Agents Are Taking Over Enterprise Work, But Most Companies Are Still Behind
Talking to your phone has long been a convenient alternative to typing. With the Pixel 11, Google is extending that option to people who communicate through Sign Language (ASL).
The viral story of an Australian entrepreneur using ChatGPT and other AI tools to create a personalized mRNA vaccine for his dog has now turned into Gamgee, a YC-backed startup developing personalized cancer vaccines for dogs.
Apple seeks publisher deals to give Siri AI better access to current events
Google updates Pixel lineup with new phones, foldable, watch
Taiwan’s Foxconn reports 35% rise in Q2 profit on AI demand, beats forecasts
Lovable confirms new $13.3B valuation, raises another $400M

The Never Ending Cycle of AI

The AI race is becoming a self-reinforcing loop. Models are getting smarter, cheaper, and more autonomous at the same time.

The recently released, xAI’s Grok 4.6 is pushing toward long-running agents that can turn ideas into working products and even patch vulnerabilities; The official version of DeepSeek V4-Pro is driving the cost of intelligence down to just $0.87 per 1 million output tokens while reportedly beating Anthropic’s Opus 4.8 on several agentic benchmarks; and Qwen3.8, with a massive 2.4-trillion-parameter architecture, is improving coding, reasoning, and long-horizon task execution.

These advances reinforce one another: better models enable more capable agents; agents generate more AI usage; greater usage drives demand for cheaper inference; and cheaper inference makes increasingly powerful models available to more developers. The result is a cycle that appears far from finished—intelligence improves → costs fall → adoption rises → usage explodes → models improve again.

The debate over open vs. closed AI

The debate over open vs. closed AI is becoming less about technology and more about who controls the future of intelligence.

At an AI4 conference, AI pioneers Geoffrey Hinton, Fei-Fei Li, and Andrew Ng all argued that AI should remain meaningfully open, but for different reasons. Ng warned that allowing a few companies to become AI gatekeepers could slow innovation and give them enormous influence over what billions of people can access.

Hinton agreed that open AI has benefits but warned that open-weight models are fundamentally different from open-source software because powerful models can be cheaply adapted for cyberattacks and other harmful uses, and, in his view, that battle is already lost. Li argued for a middle ground, comparing AI to nuclear science: knowledge can remain open while the most dangerous capabilities are regulated. The stakes are also geopolitical: Ng warned that cheaper Chinese open models could spread across developing markets and become a form of soft power, shaping how billions of people interact with ideas about freedom, democracy, and human rights.

Meanwhile, the U.S. government is moving toward bringing open-weight AI under the same safety framework as closed frontier models, potentially requiring prerelease testing once open models reach capabilities comparable to the most advanced systems, showing how quickly AI’s accelerating capabilities are forcing policymakers to rewrite the rules in real time.

The emerging consensus is therefore not “open or closed,” but how much openness each layer of AI should have, and who gets to decide.

🧰 AI Tools of The Day

AI & Tech in Crisis Response

🚁 Thermal-imaging drones - DJI Matrice 4T / Search & Rescue
DJI’s enterprise drones combine thermal imaging, AI-assisted detection, low-light cameras, and mapping for search-and-rescue operations. DJI says the Matrice 4T can be deployed rapidly and uses thermal sensing to locate people in difficult conditions.
📡 Life-detection radar — Camero-Tech Xaver Life Detection Systems
These systems use ultra-wideband radar to detect movement and breathing through walls and rubble, allowing rescuers to determine whether someone may be trapped even when they cannot see or hear them.
🛰️ Satellite imagery + AI mapping — Maxar
Maxar combines high-resolution Earth-observation imagery with AI-powered geospatial analysis to rapidly map disaster damage, infrastructure, and affected areas.

Share Yaro on AI and Tech Trends | Your Top AI Newsletter

🚀 Showcase Your Innovation in the Premier Tech and AI Newsletter (link)

Reserve a Spot

Explore sponsorship possibilities and elevate your brand's presence in the world of tech and AI. Learn more about partnering with us.

You’re a free subscriber to Yaro’s Newsletter. For the full experience, become a paying subscriber.

Subscribe for $1/Month

Disclaimer: We do not give financial advice. Everything we share is the result of our research and our opinions. Please do your own research and make conscious decisions.

Comment

Restack

871520 is your Substack verification code

noreply@blogger.com (Max) — Sat, 01 Aug 2026 15:52:11 +0000

Here's your verification code to sign in to Substack:

This code will only be valid for the next 10 minutes. If the code does not work, you can use this login verification link:

Verify email

Do not share this code with anyone. Do not forward this email. Substack employees will never ask for this code.

240

335379 is your Substack verification code

noreply@blogger.com (Max) — Sat, 01 Aug 2026 15:52:10 +0000

Here's your verification code to sign in to Substack:

This code will only be valid for the next 10 minutes. If the code does not work, you can use this login verification link:

Verify email

Do not share this code with anyone. Do not forward this email. Substack employees will never ask for this code.

310

801562 is your Substack verification code

noreply@blogger.com (Max) — Sat, 01 Aug 2026 15:31:11 +0000

Here's your verification code to sign in to Substack:

This code will only be valid for the next 10 minutes. If the code does not work, you can use this login verification link:

Verify email

Do not share this code with anyone. Do not forward this email. Substack employees will never ask for this code.

733

New privacy settings for Search services and Google Play

noreply@blogger.com (Max) — Thu, 30 Jul 2026 09:52:48 +0000

conan00maxcell@gmail.com

Hello Conan,

We're updating our settings to give you even more control over saved history and personalised recommendations across Google Search services and Google Play. Search services include Search, Maps, Shopping, Hotels, Flights, Translate and News. You'll see this change in your Google Account in the next few days.

What's changing

Previously, saving history and personalisation were managed by Web & App Activity. Going forwards, you can better tailor your Search services experience using your new Search services history and Personalised Recommendations settings. These settings let you revisit your past searches and decide if you want your experience to be personalised. As people increasingly search in new ways, like searching a photo with Lens, Search services history now includes media from your interactions, which you can stop saving at any time. For Google Play, you'll have new Play History and Personalisation in Play settings, even if you've never used this service. You can turn these settings on or off at any time.

How this affects your account

New history and personalisation settings: Your new settings for Search services and Google Play will reflect your most recent choices for Web & App Activity and Search personalisation settings. If Web & App Activity is currently off, these new settings will stay off. After the transition, the Web & App Activity setting will be separate from Search services' and Play's personalisation and history settings. Changes to one setting won't affect the others.
Auto-delete: Your prior choice from Web & App Activity for how long your history is saved will also apply to Search services history and Play history. As always, you can change how long history is saved, choose an auto-delete period and review or delete history manually at any time.
Your media is saved when Search services history is on: Saved media includes your images, files, audio and video from your interactions with Search services to help improve your experience. For example, this lets you revisit your past visual searches with Lens or continue a Search Live conversation about a song that you heard. To support these types of interactive product experiences, Google will now save your media to your Search services history, applying robust privacy and security protections. Like your Search services history, your saved media is also used to develop and improve Google services and technologies, including AI models and safety measures. With this transition, if your Web & App Activity setting is currently turned on, this new Save media subsetting will also be turned on along with your Search services history. You can turn off your Save media subsetting at any time, and you can also delete individual pieces of media from your history.

You can manage these preferences by visiting My Activity and your Personalisation settings pages.

Learn more about your Search services settings and Google Play settings.

You've received this email to let you know about important changes to your Google Account and services.

354267 is your Substack verification code

noreply@blogger.com (Max) — Tue, 21 Jul 2026 08:27:05 +0000

Here's your verification code to sign in to Substack:

This code will only be valid for the next 10 minutes. If the code does not work, you can use this login verification link:

Verify email

Do not share this code with anyone. Do not forward this email. Substack employees will never ask for this code.

451

823104 is your Substack verification code

noreply@blogger.com (Max) — Mon, 20 Jul 2026 15:38:16 +0000

Here's your verification code to sign in to Substack:

This code will only be valid for the next 10 minutes. If the code does not work, you can use this login verification link:

Verify email

Do not share this code with anyone. Do not forward this email. Substack employees will never ask for this code.

495

845971 is your Substack verification code

noreply@blogger.com (Max) — Mon, 20 Jul 2026 14:21:36 +0000

Here's your verification code to sign in to Substack:

This code will only be valid for the next 10 minutes. If the code does not work, you can use this login verification link:

Verify email

Do not share this code with anyone. Do not forward this email. Substack employees will never ask for this code.

743

937988 is your Substack verification code

noreply@blogger.com (Max) — Mon, 20 Jul 2026 13:40:39 +0000

Here's your verification code to sign in to Substack:

This code will only be valid for the next 10 minutes. If the code does not work, you can use this login verification link:

Verify email

Do not share this code with anyone. Do not forward this email. Substack employees will never ask for this code.

343

💀Why are Tech Executives Fearing for Their Lives?

noreply@blogger.com (Max) — Thu, 16 Jul 2026 19:44:32 +0000

Forwarded this email? Subscribe here for more

💀Why are Tech Executives Fearing for Their Lives?

Plus: When the Government Stalls, Citizens Code - AI in Disaster Relief

Yaro Celis

Jul 16

READ IN APP

Technology can be great at solving problems, and apps that help find missing persons due to the earthquake in Venezuela are a testament to this. When the government in this South American nation stalled. coders vibe coded apps to help families find their loved ones. We will share about these apps and how it has worked. But not all technology benefits everyone, and AI has increased the financial gaps in many communities and added to labor discrepancies. Many of us fear losing our jobs due to being replaced by this tech, while others are infuriated that is only enriching the already wealthy. Some of these frustrated individuals have taken it into their own hands to get justice, which has executives shaking. But the advancements do not stop, and Mira Murati and her company, Thinking Machines, have just unveiled an amazing open model, which is rare for a US company, that has everyone talking about it. Let’s dive in and explore what all of this means. Stay curious, and have a great weekend!

📰 AI News and Trends
Why are Tech Executives Fearing for Their Lives?
What is this New Open-Weights Model by Thinking Machines?
🧰 AI Tools - AI & Tech in Crisis Response
When the Government Stalls, Citizens Code - AI in Disaster Relief
📚 Learning Corner

📰 AI News and Trends

ASML has room to raise prices, and they will.
Anthropic moves closer to mega-IPO as bankers line up investor meetings
How OpenAI’s Sol Finally Learned Design Taste
Apple has received key approval in China that paves the way for the release of its Apple Intelligence on-device AI service in the country
DeepSeek’s Annualized Revenue Nears $500 Million, Boosting Fundraising and IPO Plans
Elon Musk quietly buys a $1 billion gas turbine company to power Grok
Hack Reveals Suno AI Music Generator Scraped YouTube, Deezer, and Genius
Moroccan intelligence insider reveals widespread use of Pegasus hacking software
Nvidia Releases New Robotics AI Model

A New Open-Weights Model by Thinking Machines

Thinking Machines has just released Inkling, a powerful new 975B-parameter Mixture-of-Experts foundation model that natively reasons across text, images, and audio.

Designed as a broad, balanced generalist rather than a narrow benchmark-chaser, Inkling excels at agentic coding, tool use, and controllable thinking effort, allowing developers to balance performance with cost and latency. Because its full weights are open, it serves as an excellent base for customization. You can start experimenting with it right away by chatting with the model in the new Inkling Playground on the Tinker console, or you can begin fine-tuning it for your specific use cases directly on the Tinker platform today.

When the Government Stalls, Citizens Code - AI in Disaster Relief

After twin earthquakes devastated northern Venezuela last month, the state’s response was agonizingly slow, but citizen developers stepped into the void. Armed with AI tools like Claude and Replit, volunteers from the Venezuelan diaspora rapidly spun up emergency response platforms in mere hours. These grassroots tools included “Desaparecidos Terremoto Venezuela”, which used AI and donated facial recognition software to process over 30,000 missing-person reports in two days, and “Ayuda en Camino”, a logistics hub built in four hours that used a WhatsApp assistant to coordinate donations in low-connectivity areas.

While experts warn that these AI-driven, citizen-led initiatives require robust data privacy measures and cannot permanently replace state accountability, they have undeniably democratized crisis response, proving that when institutions fail, communities can use AI to save lives.

🧰 AI Tools of The Day

AI & Tech in Crisis Response

If the Venezuela story caught your eye, here are 4 established apps and platforms that use AI and crowdsourced tech to coordinate disaster relief and find missing persons globally:

1. AIDR (Artificial Intelligence for Disaster Response) - A free and open platform developed by the Qatar Computing Research Institute. AIDR uses machine learning to filter and classify massive volumes of social media messages during emergencies, helping humanitarian organizations pinpoint exactly where aid is needed in real-time.

2. Ushahidi - Originally developed to map reports of violence in Kenya, Ushahidi is now a gold-standard open-source platform used in over 30 countries. It crowdsources crisis information, allowing citizens to report incidents via SMS, email, or web, and maps the data to empower rapid community response during natural disasters.

3. Google Person Finder - An essential web application activated during major crises. It serves as an open-source registry and message board for survivors, families, and emergency responders to post and search for the status of missing loved ones.

4. Zello - While not strictly AI, Zello is a vital disaster-response tool. It turns smartphones into walkie-talkies (push-to-talk) and operates effectively on low-bandwidth networks. It is widely used by volunteer rescue fleets (like the “Cajun Navy” during hurricanes) to coordinate real-time logistics when traditional communications infrastructure fails.

Share Yaro on AI and Tech Trends | Your Top AI Newsletter

Why are Tech Executives Fearing for Their Lives?

Instagram

A wave of violent threats, and at least one serious real-world attack, is targeting AI company executives and their properties. The most alarming incident was an attempted firebombing of OpenAI CEO Sam Altman’s home in April 2026, carried out by a Texas man who was later charged with attempted murder and attempted arson. Police found a manifesto on him explicitly calling for the killing of AI CEOs and investors. Around the same time, a man tailgated his way into Anthropic’s lobby, claiming an executive “was going to be killed.” A third man threatened to show up at Anthropic’s office with a pistol over a billing dispute. According to the threat intelligence firm Liferaft, the volume of digital threats targeting AI executives and data centers grew sevenfold between late February and May 2026.

By Semafor: Share of respondents that say the use of AI makes them feel concerned.

Why Are People Threatening Them

The anger is rooted in a deep and growing fear that AI is dismantling livelihoods and concentrating power in the hands of a tiny elite, which is true. Three overlapping grievances are driving the backlash:

Job loss. Companies are openly attributing layoffs to AI-driven efficiencies. When Meta announced ~1,400 layoffs in Washington state during an AI pivot, online commenters called for Zuckerberg’s yacht to be sunk or blown up. A laid-off Pinterest designer captured the sentiment bluntly: “You can’t go back to serfdom. It really feels like the people in power want to be kings.”
Economic anxiety. Ordinary people blame AI companies for rising rents, high energy costs, and a broader sense that the technology is being imposed on them without consent or recourse.
Powerlessness and alienation. Palantir CEO Alex Karp acknowledged at a conference that when people are told “your job is going to disappear,” they “go for the pitchfork.” A Quinnipiac survey from March 2026 found that Americans concerned about AI outnumber those who aren’t by more than 4 to 1, and 55% believe AI is doing more harm than good.

How Executives Are Protecting Themselves

Across the S&P 500 tech sector, the share of companies disclosing executive protection spending jumped from 26.8% in 2021 to 38.1% in 2025. Specific measures include: traveling with armed guards (though tech execs prefer low-profile, “slender” security over conspicuous bodyguards), running 24/7 security operations (Anthropic has done so since 2024), tracking “persons of interest” to catch escalation patterns early, and discouraging employees from wearing company logos in public to avoid being targeted. Some executives have also gone quieter on AI’s risks to jobs, pivoting their public messaging toward its potential benefits.

The main issue is how the industry that is simultaneously accelerating its most disruptive technology and grappling with the social fallout it is generating, is dealing with the situation. Karp put it starkly: “None of us are going to make any money when the country blows up.” The tension between the enormous financial incentives pushing AI forward and the mounting public fury at its human costs is, for now, unresolved — and the executives building these systems are increasingly living with that contradiction behind armed guards.

📚 Learning Corner

🚀 Showcase Your Innovation in the Premier Tech and AI Newsletter (link)

Reserve a Spot

Explore sponsorship possibilities and elevate your brand's presence in the world of tech and AI. Learn more about partnering with us.

You’re a free subscriber to Yaro’s Newsletter. For the full experience, become a paying subscriber.

Subscribe for $1/Month

Disclaimer: We do not give financial advice. Everything we share is the result of our research and our opinions. Please do your own research and make conscious decisions.

Comment

Restack

🏦 The Fed Is Preparing for the AI Economy

noreply@blogger.com (Max) — Fri, 10 Jul 2026 23:23:27 +0000

Forwarded this email? Subscribe here for more

🏦 The Fed Is Preparing for the AI Economy

Plus: Claude Cowork has a lot of competition. How will this change the game?

Yaro Celis

Jul 10

READ IN APP

Every AI company wants a piece of the cowork market. Anthropic has been at the forefront of general-purpose productivity agents with Claude Cowork, and now everyone seems to want a slice of the pie. OpenAI has launched ChatGPT Work, and Cursor, now owned by SpaceX, is reportedly developing “Sand ”its own version.

AI is rapidly changing the workplace, and these productivity agents are helping employees accomplish more in less time. While they promise significant gains in efficiency, they also raise an important question that if AI can do more of the work, what happens to the workers? Is this a tool to train your replacement?

But wait… Help may be on the way, or at least policymakers are paying attention. The Federal Reserve has announced a new task force to study the economic and monetary impact of AI, including its effects on productivity, jobs, and inflation. The group includes AI supporters who are against regulation and who, thanks to their investments, benefit handsomely from AI prowess. But shall see how this develops.

Let’s dive in and explore what all of this means. Stay curious, and have a great weekend!

📰 AI News and Trends
The Feds are stepping in to curb AI Monetary Damages
Claude Cowork has a competitor… Well, Maybe
Cursor is also developing a Cowork Competitor
🧰 AI Tools - Claude and ChatGPT CoWork Alternatives
📚 Resource Corner - ChatGPT Work is here

📰 AI News and Trends

Meta has released an app for making generative AI games
Apple sues OpenAI over trade secrets. Apple has lost significant talent to OpenAI as the frontier lab prepares to unveil its first hardware device this year.
OpenAI introduced a new model that listens and speaks at the same time. No more having to repeat what we already said, and hopefully stops talking when I want to say something.
SpaceXAI launched Grok 4.5 as its strongest model for coding, agentic tasks, and knowledge work, noting that it was trained alongside Cursor.
Meta AI introduces Muse Image, an image-generation model now live in Meta AI apps and Instagram Stories, offering AI image creation and editing.
Video Remix in Google Photos allows us to change backgrounds and edit existing videos and photos. Will be handy if you recently got divorced, but it may lead to some serious bullying and hacking.
Small firms are replacing tools like Salesforce and HubSpot with custom AI-built applications, cutting software costs and challenging the traditional SaaS stack
China plans to let its top AI firms buy Nvidia H200 chips
Tencent leads deal to unwind Meta’s $2bn Manus acquisition

Share Yaro on AI and Tech Trends | Your Top AI Newsletter

Claude Cowork has a competitor… Well, Maybe

OpenAI has unveiled ChatGPT Work, a major expansion of ChatGPT that moves it closer to becoming an all-in-one AI productivity platform for knowledge workers. Powered by the new GPT-5.6 model, the service combines conversational AI with advanced coding capabilities, allowing non-technical users to create documents, presentations, websites, and other business assets through simple prompts. OpenAI says the new models deliver similar performance to larger AI systems at up to one-fifth the cost, making advanced AI more accessible to businesses.

The launch intensifies OpenAI’s competition with Anthropic’s Claude Cowork and Microsoft Copilot, as the AI race shifts from chatbots to autonomous workplace agents that can complete complex, multi-step tasks with minimal supervision. ChatGPT Work will initially be available to Pro, Enterprise, and Education customers before expanding to Plus and Business users, alongside new features including a desktop app and hosted website publishing. AI companies are racing to build the “super app” for work, where a single AI assistant can replace multiple productivity tools while dramatically reducing the time and cost of everyday knowledge work.

Additionally, OpenAI launched GPT-5.6, featuring models Sol, Terra, and Luna, with Sol leading in intelligence and efficiency for coding, cybersecurity, and science. The models excel with fewer tokens at lower costs, with Sol notably outperforming competitors like Claude Fable 5. Enhanced capabilities include multi-agent parallel processing and improved design judgment, advancing cybersecurity and scientific research while maintaining strong safety measures.

But that’s not it…

Cursor is also developing a Cowork Competitor

Cursor is undertaking a major strategic pivot from an AI-powered code editor to a general-purpose productivity agent, moving into direct competition with established tools like Anthropic’s Claude Cowork and OpenAI’s ChatGPT Work. Internally referred to as “Sand,” Cursor’s upcoming agent is designed to manage everyday office tasks, such as handling emails, organizing spreadsheets, and processing text messages, signaling a significant shift to capture the broader knowledge worker market beyond software developers. This move reflects a broader industry trend where agentic AI is evolving from passive chatbots into autonomous systems that can execute multi-step workflows across a user’s desktop and file system . The expansion comes at a pivotal moment for Cursor, following reports that SpaceX has agreed to acquire its parent company, Anysphere, for $60 billion in an all-stock transaction.

The competitive landscape for these autonomous workplace agents is rapidly intensifying as companies race to solve the “last mile” of knowledge work. Anthropic’s Claude Cowork, which launched earlier in 2026, already operates as a desktop agent capable of reading local files, coordinating parallel sub-agents, and delivering finished work across applications . While Anthropic built Claude as a general-purpose model first and then layered productivity features on top, Cursor is attempting the reverse—leveraging its deep expertise in a specific vertical (coding) to expand outward into general office automation . With the AI agent market projected to exceed $10.8 billion in 2026 , Cursor’s evolution from “write code for me” to “handle work for me” positions it to capture a massive segment of everyday business operations, provided it can successfully translate its developer-focused precision into reliable, general-purpose enterprise execution.

🧰 AI Tools of The Day

CoWork Alternatives

Claude Cowork and now ChatGPT CoWork are getting all the attention, but there are alternatives, especially when you run out of credits on those.

Microsoft 365 Copilot / GitHub Copilot: Great for enterprise settings, integrating multi-model agent technologies directly into your office apps and coding suites.
Gumloop: A node-based, cloud-first platform that is LLM-agnostic, meaning you can plug in Claude, GPT, or Gemini depending on the task at hand. [1, 2]
Eigent: A powerful open-source workspace alternative featuring hundreds of tools (MCP) and multi-agent capabilities. [1]
eesel AI: Specialized for businesses, this acts like a new human hire that you train on your team’s chat history and help desk tools (like Zendesk) to automate customer support. [1]

The Feds are stepping in to Curb AI Monetary Damages

The Federal Reserve has launched five high-level task forces to review and modernize its monetary policy, signaling that the central bank is preparing for an economy increasingly shaped by AI, shifting labor markets, and new financial realities. Chairman Kevin Warsh said the Fed remains committed to its dual mandate of price stability and maximum employment. Still, he acknowledged that today’s economy has changed dramatically and that the institution’s tools, data, and policy framework must evolve with it.

The task forces will focus on communications, balance sheet policy, economic data, productivity and jobs, and inflation frameworks, bringing together leading economists, former central bankers, business executives, and technology leaders. Notably, the Productivity and Jobs group is composed of very AI-forward individuals who champion no AI and tech regulation, and a group that has and will handsomely benefit financially from AI adoption. The group includes Marc Andreessen, Charles Jones, and Microsoft’s Asha Sharma, highlighting the Fed’s growing interest in understanding how artificial intelligence and other general-purpose technologies could reshape productivity, employment, and future interest rate decisions. The initiative suggests that AI is no longer viewed solely as a technology trend; it is becoming a key factor in how policymakers assess the future of the U.S. economy.

📚 Learning Corner

ChatGPT Work is here

OpenAI will not let Claude take the reins in the AI space, and they have just released ChatGPT Work. Check the tutorial above to get acquainted with it, and please let us know how it compares with Claude Work.

🚀 Showcase Your Innovation in the Premier Tech and AI Newsletter (link)

Reserve a Spot

Explore sponsorship possibilities and elevate your brand's presence in the world of tech and AI. Learn more about partnering with us.

You’re a free subscriber to Yaro’s Newsletter. For the full experience, become a paying subscriber.

Subscribe for $1/Month

Disclaimer: We do not give financial advice. Everything we share is the result of our research and our opinions. Please do your own research and make conscious decisions.

Comment

Restack