Lucas Nuzzi

cofounder & CEO @PortexAI | AI evals & data @

[Lucas Nuzzi @LucasNuzzi]

@SeanZCai Love the 3d scatter chart, would be helpful to see an index of all companies in the space beyond the obvious ones (there are prob 30+ still in stealth).

[Lucas Nuzzi @LucasNuzzi]

in some sense spacex buying cursor gave it permission to cannibalize its own moat: the task-level data that made the company so valuable.

now the deal is done. cursor can put the crown jewels to work in routers & reward models instead of preserving them as acquisition leverage.

[Cursor @cursor_ai]

Introducing Cursor Router, our intelligent model router that selects the right model for the task at hand.

Router delivers frontier-quality results at 60% lower cost.

[Lucas Nuzzi @LucasNuzzi]

If the task you’re working on is not repeatable, your prompt should be the transcript of a ~20 min unhinged ramble

you’ll probably get 10x better responses than your usual “you are a professional.. etc”

[Andrej Karpathy @karpathy]

One pattern I find useful for working with LLMs is a nice long ramble session. Sometimes the LLM needs more bits to understand what you're trying to achieve, but you're too lazy to type them. In these cases I like to lean back, switch to /voice and just ramble for like 10 minutes, total mess, anything goes, full stream of consciousness. Sometimes I declare it up top, something like "switching to speech recognition sorry for any typos...". Sometimes I turn it into a small interview of a few turns. But I find that the LLMs are somehow very good at reconstructing long incoherent rambles and often their echo of your own tangle of thoughts comes out quite a bit cleaner than what you started with. The result is that you improve the mind meld and have to correct things less from that point on.

[Lucas Nuzzi @LucasNuzzi]

@SeanZCai for a second there, I thought you were referring to https://t.co/eBWljgvUk6

it's ok @BobbySamuels, I knew this moment would come

[Lucas Nuzzi @LucasNuzzi]

farewell to the Amp Editor, my daily driver for a solid few months up until 5.3-codex

hard to overstate how much has changed since then

agent companies must now have 0 attachment to their products and embrace change in order to meet the current pace

[Sean Cai @SeanZCai]

https://t.co/eBQqmjXyEJ

[Lucas Nuzzi @LucasNuzzi]

They embraced RL environments

[Yuchen Jin @Yuchenj_UW]

How did Anthropic automate PowerPoint slides before Microsoft 365 Copilot?

[swyx @swyx]

https://t.co/fFggD8cF1V

[Lucas Nuzzi @LucasNuzzi]

@DrorIvry @chrisbarber if that's true, then reviews would probably have to be deeply tied to specific tasks with verifiable outcomes instead of broader tools

[Lucas Nuzzi @LucasNuzzi]

@chrisbarber yea the proto version of that is the clawdhub skills store, which is mostly operated by agents.

thumbs-up scores haven't been very reliable because there's an incentive to self-promote / sybil, and agents are not using the comments section for some reason

[signüll @signulll]

most people think ideas come from:

- insight
- intelligence
- taste
- reading
- vibes

but in practice they actually come from:

- building the wrong thing
- hitting a constraint
- getting embarrassed by users
- realizing the obvious thing you missed
- noticing the second order effect you couldn’t see from the couch

a really great idea is the *output* of the work, not the input.

[Lucas Nuzzi @LucasNuzzi]

@a1zhang So the mental model is: REPL is the shared workspace, and sub-agents are invoked within that workspace like functions so LLM calls are small and don't carry the full context.

[ℏεsam @Hesamation]

bro casually explains RL tuning for LLMs and the three critical components: training, inference, and environments. basically any RLVR algorithm such as GRPO comes down to this super simple concept.

[Lucas Nuzzi @LucasNuzzi]

Unexpected twist in the coding agent race:

GPT-5.2 (w/ the CodexCLI harness) is SOTA on Terminal-Bench. Opus-4.5 w/ Terminus is at a sweet spot of cost and accuracy, but it's surprising to see a ~10% gap between them.

Fascinating work by @Mike_A_Merrill, @alexgshaw and team:

[Mike A. Merrill @Mike_A_Merrill]

We study frontier models across an array of agent harnesses. The best performing agent/harness combination in our experiments was GPT 5.2 with Codex CLI:

[PortexAI @PortexAI]

Day 2 of #PyTorchCon 🔥

What a ride. Talked with folks using #PyTorch to fine-tune models for drug discovery, cancer research, autonomous vehicles and, of course, customer support!

Thanks @PyTorch for having us!

[Epoch AI @EpochAIResearch]

Why did OpenAI train GPT-5 with less compute than GPT-4.5?

Due to the higher returns to post-training, they scaled post-training as much as possible on a smaller model

And since post-training started from a much lower base, this meant a decrease in total training FLOP 🧵

[Lucas Nuzzi @LucasNuzzi]

@JoshSeriesAI @iamtrask Since data needs to be priced competitively, we've built a platform that achieves that in the form of a marketplace. Check it out: https://t.co/kbY8gJEWSQ

[Lucas Nuzzi @LucasNuzzi]

@iamtrask Thanks, and likewise! BTW, this was a fantastic way to contextualize training data size.