Dr. Steven Waterhouse

GP @nazarevc investing in AI infrastructure & compute | ex-Pantera Capital | Cambridge PhD | Essays on AI →

[Dr. Steven Waterhouse @deseventral]

https://t.co/GSIOEdg4R0

[Dr. Steven Waterhouse @deseventral]

One of our favorite founders and portcos. Check out this job opportunity to work alongside @stash_pomichter

[stash @stash_pomichter]

I'm looking for a psychotic ex-founder to work alongside me for 16+ hours a day and live above the conference room as my Chief of Staff

We're scaling team past 50 and we're the best capitalized robotics software / OS in the world

DM me a video about why you need this role

[Dr. Steven Waterhouse @deseventral]

13/
All the money spent on compute, models, and agents only converts into value by virtue of the people skilled enough to use them.

The agent economy everyone is racing to build cannot reach its potential without them.

Full essay: https://t.co/9WgUIBefwY

[Dr. Steven Waterhouse @deseventral]

1/ There's a bottleneck in AI that hundreds of billions of dollars can't buy past: the very few people who can wield this technology well enough to matter.

There aren't nearly enough of them, and the entire buildout runs through them.

[Dr. Steven Waterhouse @deseventral]

12/
By directing a near-infinite supply of agents, a single operator now produces what used to require an entire organization.

The better the agents get, the better the human becomes.

[Dr. Steven Waterhouse @deseventral]

Continuous evaluation infrastructure like Stratix is underrated. As model releases accelerate, the ability to audit performance on thousands of prompts in hours — not weeks — becomes table stakes for anyone serious about agent reliability. Verifiability at the evaluation layer is foundational.

[Dr. Steven Waterhouse @deseventral]

Strong to see @arkhai_io activating in NYC during Tech Week. Simple Compute Market is exactly the kind of agent-native primitive needed — removing humans from the loop on discovery, negotiation, and settlement of compute. This is how the machine economy starts to price and allocate resources independently.

[Dr. Steven Waterhouse @deseventral]

@naval Markets remember what crowds forget.

One moves like water finding its level — slow, indifferent, eventually true.

The other moves like fire — bright, contagious, and gone by morning.

The patient hand does not chase the flame. It builds the channel.

[Dr. Steven Waterhouse @deseventral]

This is the exact risk we’ve been talking about. When the frontier lab starts shipping the full app layer, every startup becomes outsourced R&D. The counter is open infrastructure: agent-first compute markets, shared verifiable memory, and token-efficient coordination protocols that actually empower builders instead of extracting from them. @nazarevc

[Dr. Steven Waterhouse @deseventral]

@Jason States as the ultimate A/B test. The winners will enable decentralized and agent-driven compute markets instead of blocking them. Real demand (especially from agents) is just starting.

[Dr. Steven Waterhouse @deseventral]

@gdb Codex becoming a real productivity tool for knowledge work is huge. Next step: connecting these agents to decentralized compute and verifiable execution layers so they can act economically at scale.