The real win from being able to train a 1T parameter model on a “shoe string” budget isn’t the cost savings. It’s the efficiency gain that lets you move faster and increase your iteration speed. Pay attention to the slope. Since I can remember, the best deep learning models

Intelligence Isn’t Enough: Why Energy & Compute Decide the AGI Race – my conversation with @eisokant, co-CEO of @poolsideai 00:00 - Cold open – “Intelligence becomes a commodity” 00:23 - Host intro – Rumored $2B round; Project Horizon & RL2L 01:19 - Why Poolside exists amid

We are building together with our customers. Our Forward Deployed Research Engineers are closing the gap between model + agent capabilities and customer's hard problems. We're hiring: https://t.co/n1HMSRUKYb

Heading over to @mattturck's office to record a podcast, almost 2 years since we recorded the first podcast we did for Poolside. Any questions/topics you'd like us to cover? On the episode 2 years ago, it was highly contrarian to describe our approach to Reinforcement Learning.

Really excited to be partnering with Alex & Redpanda! Expanding our platform and opening up our models and agents to +300 enterprise data sources + a highly capable query engine is a huge unlock for the capabilities we can deliver to our customers. And man are they great to

Great work! Very bullish on this direction. Almost everything we do in foundation models is either improving compute efficiency or improving data. Allowing for adaptive computation can be one of the bigger unlocks for improving model capabilities within a compute budget.

Can someone please create this TV show:

some thoughts on @poolsideai's project horizon announcement last week: building a 2gw ai factory "This level of vertical integration - from model and agent development through to power generation and deployment - gives poolside a durable advantage as AI systems become

Would love to connect with anyone who's impacted, and is looking to join a small, but well-resourced team to push to the frontier and beyond. We have one of the highest ratios of GPU resources per researcher. No politics or siloes.

Will be a fun conversation! Should be on at 1pm PST

We believe that to compete at the frontier, you have to own the full stack: from dirt to intelligence. Today we’re announcing two major unlocks for our mission to AGI: 1. We're partnering with @CoreWeave and have 40,000+ NVIDIA GB300s secured. First capacity comes online

I’ve known Nathan for almost a decade and for the last 8 years have seen him and his team treat this as a labor of love. Thank you for creating one of the most comprehensive reports on our industry!

Thank you 😆

Ari is one of my favorite people in our field. Worth listening to him. When we started poolside, one of our guiding research beliefs was "not all tokens are equal".

Founders WhatsApp when you're both vibe checking your latest model

What if training a foundation model felt less like chaos… and more like running a finely-tuned factory? 🏭 @poolsideai Model Factory shows how. A quick tour 👇

In the limit, evaluations are the ~only thing that matters. When models are self-improving, and every metric can be hill climbed, picking the metric becomes the most important thing. Evals will shift from being "writing unit tests" for research to being the *main thing*

We've not been very public about our progress on model building, but I fully believe poolside will be the next lab joining the frontier. We're now sharing a bit more how we're doing this, with a systems-first approach we're taking with our model factory.