Chris Messina

$64M Seed Round for a Self-Improving AI Mathematician

  1. Surely there is no bubble.

  2. And, can you smell the AGI now?

Axiom's mission is to build a self-improving superintelligent reasoner, starting with an AI mathematician, and for that, Carina Hong has raised $65M in seed funding.

I snark, but this is one reason why VC doesn't make sense like it did in the SAAS era... it's not just about winner take all dynamics, but about putting a dent in the information theory and our grasp of consciousness.

Another to-do app just ain't gunna cut it when you could fund a to-do app that checks off its own tasks. ✅

278 views

Add a comment

Replies

Best
Marcello Cultrera

A similar approach is attempted by Mira Murati at Thinking Machines where most researchers are either quants or mathematicians. Axiom’s approach of starting with an AI mathematician is inline with this more recent trend shifting from earlier focus on autoregressive transformers pattern recognition (LLMs/ data science engineering) to more theoretical work on structured reasoning. It's brilliant.

Sourn Rockit

A self-improving AI mathematician sounds like the start of a sci-fi novel… except it’s real life now.

Abdul Rehman

Finally, a to-do app that doesn’t guilt-trip me with notifications… it just does the tasks itself.

Sanskar Yadav

It's wild how quickly the lines between startups and scientific research are blurring.
From SaaS to self-improving mathematicians. Colossal Labs is bringing Dodo back.

This pace of progress makes it hard not to be excited (and maybe a tiny bit cautious). Feels like we’re watching the next chapter of tech being written in real time, and even we don't know what the next line is going to be

Ray Wang

To be honest, there are probably enough mathematical work from hedge funds and research universities where 64m seems a small amount to even get the junior level work outsourced.

Where is the disconnect to you? The quant team at one small hedge fund probably cost 10-20 mil.

Dawid Baranowski

Maybe not quite AGI, but I'm personally very bullish - and excited- on AI applications in hard science. Will be interesting to watch what happens to IP when more and more deep knowledge discovery / consciousness envelope pushing is done by VC backed enterprises.

Does start to feel more like sci-fi everyday.