Seth DeGroot

Well, perhaps this shall reach Andilar Although I know not how it can For once again, he's hurled his wind Upon the silver prow But if it should, my words are these "Arise, young men, fine ships to build And set them north for Valinor 'Neath standards proud as fire"

Tallow

Bitcoin is the largest idle asset in the world. I backed Zest Protocol because they're the team putting BTC to work. BTC-backed lending is a trillion dollar market and @ZestProtocol has the track record to win it.

I felt a great disturbance in the Force... as if millions of PropAMM and meme-pad voices suddenly cried out in terror and were silenced.

@andrewchen @wkoszek x.com/ishaan_jaff/st…

@wkoszek Yeah seems like the best you can do rn with coding harnesses is to point it at a liteLLM instance, and then configure the routing there - since liteLLM can also send to Claude and GPT, in addition to local

How soon before a real % of LLM queries are done via local AI models running webGPU in-browser, and are never sent to the SOTA model in the cloud? Couple things that might drive this: - you don’t need a frontier model for everything. A very large % of LLM queries are simple, google like queries. Easily handled - local models are getting really good, and getting better - a lot of consumer hardware (particular Apple!) can already run good models pretty well. Newish mac laptop running qwen 3.6 35b MoE LLM at great speeds. Hardware is going to get even better in this direction - there’s def some use cases where people will care about privacy. Health, financial, adult stuff etc. - nice part about browser/webGPU is that there’s no install. It’ll just work. And alleviate compute costs Of course the tension for this is that we’re just going to build a shitload of compute in the world, and tokens will get cheaper over time. Yet it seems like the demand is also so crazy that unlocking a bunch of local supply will be worth it too

We're open sourcing LiteLLM Agent Platform Run Claude Code, Codex, Hermes or any coding agent in isolated K8s sandboxes The agent never sees your real API keys. Here's how 👇

@brim006 View from the summit

Sure, Everest is the highest mountain in the world by elevation, but it barely cracks the top 10 for actual height from base to peak. In that category, the USA wears yet another global crown thanks to Denali. 🇺🇸 And if we include underwater mountains, USA of course takes that crown too thanks to Mauna Kea. 🇺🇸 It has always been odd and surprising to me how rarely this gets mentioned. So just remember, the USA is the global king of everything, even mountains. Tallest Mountains by Base-to-Peak Rise: 1. Denali (Alaska, USA) — ~18,000 ft (5,500 m)
Rises from a 2,000-ft plateau — the largest vertical relief on land anywhere. 2. Mount St. Elias (Alaska/Yukon) — ~18,000 ft (5,486 m)
Soars almost entirely from near sea level (coastal tidewater) to its icy summit. 3. Rakaposhi (Karakoram, Pakistan) — ~17,060 ft (5,200 m) rises in a mere 7 miles. 4. Nanga Parbat (Himalayas, Pakistan) — ~15,090 ft (4,600 m)
Home to the massive Rupal Face, one of the largest continuous walls of rock and ice on Earth. 5. Mount Kilimanjaro (Tanzania) — ~15,100 ft (4,600 m)
The world’s tallest free-standing mountain, rising dramatically from the surrounding plains. 6. Mount Fairweather (Alaska/B.C.) — ~15,000 ft (4,570 m)
Abruptly rises from the maritime lowlands of Glacier Bay. 7. Pico de Orizaba (Mexico) — ~14,500 ft (4,420 m)
A massive volcanic cone towering over high surrounding valleys. 8. Mount Gongga (Minya Konka, Sichuan, China) — ~13,500 ft (4,115 m)
Framed by deep river gorges plunging into the Dadu River valley. 9. Mount Everest (Nepal/Tibet) — ~12,000 ft (3,650 m)
Highest point on Earth, but its base sits on the elevated Tibetan Plateau. 10. Mount Rainier (Washington, USA) — ~11,400 ft (3,475 m)
Largest localized vertical relief in the contiguous United States.

$CELC announces updates to our clinical development plan including expansion of our Phase 3 VIKTORIA-2 clinical trial to include a second study evaluating gedatolisib as 1L treatment in patients with endocrine-sensitive HR+/HER2- advanced breast cancer. bit.ly/3P5iWQG

We spent the past 3 years helping our partners find, deploy, and orchestrate compute across 30+ clouds. With the unprecedented demand on the market right now and the largest shortage of compute and power we’ve seen to date, we’re excited to launch Shadeform Capacity Solutions (SCS). SCS starts with Shadeform delivering a full market wide assessment of available inventory and buildouts coming online in 2026 and 2027. If there is availability for your deployment, great - we work hard to deliver a reliable, scalable cluster on time and ready for your workload. If there isn’t a planned deployment that matches, we build. We work with our real estate, datacenter, OEM, and cloud partners to deliver a full packaged solution with compute installed and ready to go for your custom needs, so you can predictably scale. We're not going vertically integrated. The way to think about it is that we're the developer and PM - we put together the assets, then deliver them to our cloud partners to extend their portfolio and manage and operate a new cluster for you. In a real estate model, our partners are the owner and property manager; we're the developer; our customers are the tenant.

“Prolonged calculator use may make it harder to reason numerically and do mental arithmetic,” per the Economist, circa 1975

Today we are launching the @LiteLLM Agent Platform - a simple, self-hosted infrastructure platform for running multiple agents in production. This is great for AI platform teams looking to enable their developers to ship agents without giving up control of their infra.

just ~six months ago so many ppl thought everything was bubble… e.g. too much compute, too much capex, & demand that couldn’t possibly absorb the buildout. but it turns out the ceiling on demand for intelligence is literally nowhere in sight. i mean who could’ve ever seen that a smart entity available on tap is actually insanely useful for ~every facet of human life & even for things we haven’t even thought of that are possible now???

Excited to announce @ZestProtocol Bitcoin Collateral Vaults at Draper Summit today. BTC in. Stablecoins out. BTC stays on L1 🟠

@jrosenfeld13 @charliermarsh RATE_LIMIT_PAUSE

Having seen ~3-4 tweets on the matter, I'm 99.99% confident that OpenAI will build a GitHub competitor powered by @charliermarsh magic

Art deco data centers. Who's building this?

not enough people are emotionally prepared for if it’s not a bubble

🥩

Sourcing GPUs has turned into a part-time job. Endlessly pinging contacts. Filling out the same form on 5 provider sites. Waiting days for pricing. Shadeform lets you skip that. Just tell us what you need, and we'll come back with options from 30+ clouds in under 24 hours. shadeform.com/features/reser…

My view is the AI boom will only accelerate and is a once in a lifetime transformation This is orthogonal to whether many AI companies should exit in next 12-18 months, as some may lack durability vs labs, new entrants, or weird market shifts businessinsider.com/ai-startup-fou…

Soon.