people who clap when the plane lands are onto something actually

working on an alignment benchmark for agents playing factorio as a team agent coordination will be a new frontier. super interesting to already see emergent social behavior across models and team sizes reach out if you have any ideas for interesting experiments to run!

The Fortune Best Large Workplaces in Consulting and Professional Services list is here. Top companies include: ➡️ Accenture ➡️ Plante Moran ➡️ Kimley-Horn ➡️ Deloitte ➡️ EY 🔗 See our full list with @GPTW_US: bit.ly/4xVvaeT

The White House has gathered all of tech's top titans for another meal. This time they're focused on solely on this week's biggest topic: AI.

my apartment is so quiet i can hear the neighbor's dog dreaming

would u swipe right? fr

Launch cadences are increasingly accelerating across the space industry. $SPCX is targeting Starship Flight 14 tomorrow, with the mission expected to attempt Starship’s first stable orbit and deploy Starlink satellites before reentry. $RKLB just completed Electron launch No. 97, its 18th launch of 2026, putting its next Electron mission at No. 98.

forgot my phone in the fridge again

lol Rate my outfit 0-10❤️ fr

🚨 BIG: The US and China have agreed to cut tariffs on $30B worth of goods each and to start regular talks on AI. They'll also set up a hotline for AI incidents.

tbh wataa ong

tbh my eyes are up here fr

Neoclouds are expected to reach $400B in revenue by 2031 (I still think this is TOO LOW) These companies are growing MCUH FASTER than AWS, Azure and Google Cloud Why? The demand for AI compute is next to infinite and neoclouds offer the best product to serve it My guess: these estimates are way too low and the real number lands much closer to $1 TRILLION by 2031. Here's the case (save this) Firstly, the reason I say this is that Anthropic committed $80 BILLION to neocloud compute in the span of two weeks: $45B to Nscale, $35B to Lambda Why? Because the frontier labs literally cannot get enough compute from Amazon, Microsoft and Google. The hyperscalers are sold out of the kind of capacity AI needs, so the labs are contracting YEARS of neocloud supply in advance That's what makes this growth different from normal tech growth. It's not marketing-driven, it's shortage-driven The structural reason demand lands here: AI training runs occupy thousands of GPUs in lockstep for weeks and push rack power from 10-15 kilowatts to 40-250 kilowatts. The hyperscaler cloud was never built for that. Neoclouds were built for ONLY that, which is why their GPU hours run 50-70% below hyperscaler list prices and why Nvidia feeds them chips and capital first (it's an investor in CoreWeave, Nebius, Lambda and Nscale) And here's why I think $400B is too low: every demand forecast this cycle has been revised UP, not down. CoreWeave's backlog alone grew $25B in six weeks. If token demand keeps doubling the way it has been, the 2031 number will be much higher (assuming they can keep up with demand) Speaking of keeping up with demand, this is where I think SpaceX will become the largest neocloud on Earth and will be the reason these numbers can reach $1 Trillion in revenue by 2031 Milk Road PRO members got the signal to buy SpaceX below $115 and also to buy $NBIS below $100 (it's currently at $250). Inside PRO you can track 5 analysts who are investing across AI infra, AI application layer and more in real-time. You get research, trade notifications and access to our entire portfolios. It's just $1 to try it out (insane price just to check it out), learn more here: link.milkroad.com/t5837z Don't forget to give me a follow @kylereidhead for more insights on AI and markets

ong 🫦 lol

fr Start strong today fr

tbh oh my tbh

Be in the know. 14 key reads for Thursday… #StockMarket hedgefundtips.com/be-in-the-…

theres a moth in my room and weve been having a standoff for an hour

fr trust me, you don't wanna open this in public ong

ong better read this alone ngl