TL;DR
Meta and BlackRock form a $14B joint venture, with Wall Street as landlord for Meta's AI compute
Denovo turns a one-line idea into a live business with paying customers
Synthesia lets you translate any video into five languages using your own voice and face
Over 1,000 AI lab staffers ask the US to help build a brake pedal for frontier AI
Webflow's MCP 2.0 gives AI agents brand-aware, permission-scoped control over your site
Anthropic's unreleased model found cryptographic flaws elite humans missed for years
📈 Wanna know what CEOs & Founders are reading to stay ahead in Tech & AI? It's called The CEO Report and it's trusted by 250,000+ fast growing, smart decision-making C-Suite execs!
What are you waiting for, sign up for free here →
META RENTS ITS OWN DATA CENTER FROM WALL STREET
Meta and BlackRock just formed a joint venture worth roughly $14 billion to build a 1 gigawatt AI data center campus in El Paso, Texas. BlackRock funds own 80%. Meta owns the other 20%, and leases the whole thing back once it's done.
Here's the structure. Meta puts in land and existing construction, about $2.3 billion worth. BlackRock puts in $4.9 billion cash, a chunk of it borrowed. Meta will be the only tenant when capacity comes online in 2028, running the place day to day while BlackRock's money sits underneath it.
This isn't a one-off. BlackRock just formed a similar venture with ACS called Coravel, also in Texas, and seeded another one in the UK. Wall Street isn't just writing checks for AI infrastructure anymore. It's becoming the landlord.
That matters because the AI buildout is enormous, and even Meta doesn't want the whole bill sitting on its own balance sheet. Structuring it this way lets Meta scale compute fast without owning the asset outright, and lets BlackRock collect steady returns from a tenant that isn't going anywhere.
OUR TAKE
Wall Street just became a core co-owner and lender in the AI compute arms race, not just a source of cheap capital.
This kind of structure lets Meta scale faster without loading the debt onto its own balance sheet. BlackRock gets a steady tenant and a slice of the infrastructure boom instead of just underwriting it from the sidelines.
The risk and the returns are getting spread across the financial system now, not just sitting on Big Tech's books.
Watch who becomes the landlord next. Whoever owns the buildings ends up with real leverage, even if they never touch a GPU.
QUICK HITS
AI lab staffers ask for a brake pedal, not a stop button: Over 1,000 employees from OpenAI, Anthropic, Google, and nearly a dozen other labs signed a letter asking the US to help build tools that could deliberately slow frontier AI progress. It's not a pause, it's an option, giving governments and labs a way to buy time if AI research starts accelerating faster than anyone can track.
Webflow ships MCP 2.0, and it's not just an API wrapper anymore: The new version gives AI agents brand-aware design controls, permission-scoped CMS access, and the ability to spin up entire sites on Webflow Cloud without a human touching the provisioning step. An agent can now pull your typography, colors, and component library and actually build inside your brand instead of guessing at it.
Anthropic's unreleased model just out-mathed the cryptographers: Claude Mythos Preview found a lattice flaw in post-quantum candidate HAWK that human researchers missed for years, and invented a technique that speeds up attacks on 7-round AES by up to 800x. Nothing here breaks production systems yet, but AI is now finding weaknesses elite cryptanalysts spent years missing.
You already follow the model launches, benchmarks, and breakthroughs. Now trade on what happens next. Explore real-world AI and tech markets on Kalshi. Trade $25, get up to $500.
STOP LOSING CUSTOMERS WHO DON’T SPEAK ENGLISH
Turn one video into five languages using your own voice and your own face, no re-recording required.
⏱️ 15 minutes | 🔧 Synthesia account (free tier works), one video you've already made
Why build this? Most videos never leave their home language. Translating one properly used to mean hiring a voice actor, re-recording everything, and syncing it back up, which costs real time and real money. So creators skip it and leave most of their potential audience untouched. Synthesia's video translator clones your voice and regenerates your lip movements to match, so the output looks native instead of dubbed.
Steps:
Go to Synthesia's video translator and upload your file (MP4, MOV) or drop in a YouTube link
Let it auto-detect your source language; no setup needed
Pick your target language and hit generate
While it processes, open the transcript editor to fix translations, lock brand names, or adjust pronunciation. Editing here doesn't cost credits
Choose adaptive mode (matches original pacing) or original mode (natural playback speed), test both the first time
Download the per-language MP4, grab the SRT file, or use one smart link that auto-serves each viewer their language
Expected outcome: One video becomes a library. A product demo turns into a pitch that works in five markets without hiring anyone. A training video finally makes sense to the part of your team that doesn't work in your primary language. A sales clip stops losing replies just because the buyer doesn't speak English. Same video, same voice, same face, just fluent everywhere you need it to be.
Full tutorial:
TOOL OF THE DAY
You had the idea. Denovo does everything after that.
Tell it your business idea and it builds the plan, the website, the branding, wires up Stripe, then runs your ad campaigns and starts bringing in customers. No code, no waiting around for a dev team.
It's not just a tech founder thing either. People are launching fintech, e-commerce, coaching businesses, wedding startups, all built end to end by the same AI cofounder.
Best for founders who have the idea but not the six months it usually takes to turn it into something that makes money.
