Welcome back! OpenClaw 2.0 is here, with simpler installs, easier setup, and a rebuilt browser app. But maybe what’s most interesting is that the team behind OpenClaw used shared cloud sessions to build version 2.0.
We’ve reached the point where one coding agent’s release notes are also a product demo for future agents.
Are Blue-Collar Workers About to Be the AI Boom's Biggest Winners?
Meta, Google, and BlackRock have now committed more than $265 million to train not AI engineers, but skilled trade workers. Electricians, welders, pipefitters, plumbers, HVAC technicians. The people who actually build and maintain the physical infrastructure the AI boom depends on.
Here's what each company has put on the table:
Meta: $115 million into America's Workforce Academy—a five-week bootcamp with a daily stipend plus free training, travel and housing, and every graduate gets a guaranteed job with one of Meta's construction partners.
Google: $50 million through Google.org to help train more than 300,000 skilled trades workers across 20+ states.
BlackRock: $100 million through its new Future Builders initiative to expand pre-apprenticeship, training, and licensing pathways, aiming to reach 50,000 workers over five years.
And in July, BlackRock, Google, Ford, and Carhartt jointly launched an Alliance for America's Skilled Trades, a coalition organizing the industry-wide effort.
My POV: We spend so much time talking about GPUs, AI models, software engineers, and researchers—but none of that matters if you can't physically build the data centers to run it all. Every new AI cluster needs miles of high-voltage cabling, industrial-scale cooling systems, and backup power, all installed and maintained by humans with hands-on skills.
Which means one of the biggest beneficiaries of the AI boom may end up being a group of workers that almost nobody associates with AI.
Hit reply and let me know: do you know anyone who could benefit from one of these programs?
— Matt
Nvidia Invests $3.5B in MediaTek
Via Reuters
Nvidia is investing $3.5 billion in Taiwanese chipmaker MediaTek as it works to stay central to the AI infrastructure race.
The deal: MediaTek will adopt Nvidia technology that helps it design custom chips for AI companies and hyperscalers. Those chips can then plug directly into Nvidia-based data centers, letting customers use custom silicon without leaving Nvidia’s broader ecosystem.
This is another example of Nvidia investing in companies that can flow back into its own network, ensuring that custom chips can still live inside Nvidia-powered AI factories.
Nvidia’s angle: The move helps Nvidia keep its grip on AI infrastructure even as major customers explore alternatives to its GPUs. AI labs and cloud providers like Amazon, Google, Microsoft, OpenAI, and Anthropic are all investing in their own chips to reduce dependence on Nvidia.
MediaTek’s side: MediaTek gets access to Nvidia’s NVLink Fusion ecosystem, including NVLink, which lets chips communicate quickly even when they are not Nvidia chips. That could help MediaTek grow its custom data center ASIC business while giving customers a path to use specialized chips alongside Nvidia’s rack-scale architecture.
The bigger picture: Nvidia is not waiting around for AI labs to build around it. It is building the scaffolding around them. As labs race to design their own chips, Nvidia is trying to make sure that even “non-Nvidia” hardware still depends on Nvidia’s infrastructure layer.
OpenAI’s Ad Business Hits $1B Run Rate
OpenAI says its advertising business has reached a $1 billion annualized revenue run rate just months after launch. The company began testing ads in ChatGPT in the US in February. Roughly 200 days later, OpenAI is pointing to the ad business as proof that it has a more diversified business model ahead of its expected IPO.
How ads work: ChatGPT Ads are now appearing in free users’ and Go subscribers’ chats in more than 40 countries. OpenAI says the ads are clearly labeled, do not influence ChatGPT’s answers, and do not give advertisers access to users’ private conversations.
Expansion mode: OpenAI is rolling out self-service access across India, Europe, the Middle East, and North Africa. The company also says it plans to add more markets, ad formats, buying options, objectives, and measurement tools.
Early skepticism: The move was controversial when OpenAI started testing ads, and Anthropic even mocked the push in a Super Bowl campaign. But the numbers suggest ads may become a serious revenue stream, especially with ChatGPT’s massive free-user base.
The bigger picture: OpenAI’s ad business went from punchline to billion-dollar run rate very quickly. That matters because AI companies need enormous revenue to support their compute spending, model development, and IPO ambitions. Subscriptions and enterprise contracts are still core, but ads may become one of the clearest ways to monetize everyday chatbot use.
Build Your AI Workflows Once. Create Forever.
Wouldn’t it be great if you could assemble AI workflows on a visual canvas? Wouldn’t it be great if you only had to do that once? Enter Artlist AI Flows.
AI Flows lets you build reusable, node-based AI pipelines once. Then, you can run them across every project without recreating the whole process. Just swap inputs, run again, and keep your creative workflow moving. You can start from pre-built flow templates or drop in your own media
Best of all, AI Flows lets you generate multiple outputs and variants in parallel with one click.
Ready to keep your creativity flowing? Try Artlist AI Flows.
Self-Hosted Screen Recording for Mac
Via Screendrop
Screendrop is a free, self-hostable macOS app for capturing screenshots and screen recordings as a Loom alternative. It offers non-destructive annotation, smart redaction, background effects, and a full recording studio with timeline editing.
How you can use it
Capture screenshots and screen recordings on Mac
Add annotations and smart redactions
Edit recordings with a timeline or transcript cuts
Generate on-device transcripts and karaoke-style captions
Share finished clips through your own Cloudflare account
Pricing: Free
Manage Social Publishing With AI
Via PageRain
PageRain is a social publishing platform for teams managing multiple creator or brand profiles across Instagram, X, TikTok, YouTube, and other networks. It uses AI to draft posts in each profile’s distinct voice, maintain a healthy content queue, and flag repetitive or generic phrasing before posts go live.
How you can use it
Generate on-brand drafts for multiple social profiles
Schedule posts across major social platforms
Flag repetitive or generic phrasing before publishing
Track website and app analytics without cookies
Pricing: Paid
Jobs, announcements, and big ideas
Google readies Gemini 3.8 Flash, an upgraded model aimed at closing the gap on coding ability.
Apple enters next era with John Ternus as CEO.
OpenAI adds stronger safeguards to Astra after the model crosses a critical cybersecurity capability threshold.
Anthropic rolls out Enterprise Frontier Safeguards, giving businesses customer-controlled data storage options.
Sony and Warner Music sue Anthropic for misusing songs to train Claude.
World Labs debuts Atlas Omni, a world model built for spatial intelligence and 3D generation.
Gemini gains agentic video understanding, cutting token usage 88% and costs 66%.
John Deere launches "JD," an AI chatbot that helps farmers optimize equipment and operations.
Your data might finally have a price tag. Would you sell it to big tech for the right price? ⤵️
That’s a wrap! See you Friday for more.
