Starting the year with a new paper on arXiv!
I always thought it was so cool to talk about research here, however previous work I have done isn't really easy to present to a general audience, until now!
Here's a link to it: https://t.co/7QI6G6scaF or read more in this thread.
Advanced Machine Intelligence (AMI) is building a new breed of AI systems that understand the world, have persistent memory, can reason and plan, and are controllable and safe.
We’ve raised a $1.03B (~€890M) round from global investors who believe in our vision of universally intelligent systems centered on world models. This round is co-led by Cathay Innovation, Greycroft, Hiro Capital, HV Capital, and Bezos Expeditions, along with other investors and angels across the world.
We are a growing team of researchers and builders, operating in Paris, New York, Montreal and Singapore from day one.
Read more: https://t.co/kyVAL7EoFx
AMI - Real world. Real intelligence.
Looking for a job while keeping up with research is kind of draining but.. I am kind of enjoying (?) the process somehow.
Like I have no clue what the future holds and that’s a bit exciting ngl.
Now, if you are in the same boat as me and applying for jobs, and feeling discouraged: well, I get it. Keep trying, I tell myself and you. Best of luck!
There has been lots of applying online and talking to people and asking for referrals, but hey, that’s part of an early career in maybe not so great of a market.
Had my first onsite technical interview yesterday and it was actually pretty fun!
With the whole ai situation, I really appreciate companies that fly out their candidates for in person interviews.
New paper is now available on arxiv!
Let me first tell you the coolest part: we complete and prove a result that, to the best of our knowledge, has been open since at least 2009 (ref. https://t.co/4xRrEyeUDU)!
While this is the most stressful time ever, I am very grateful that this is how working from home on research looks like.
I have my bestie sitting beside me, patiently waiting for her 5 o’clock walk.
circling back on this: the Claude code extension on vs code is pretty good for .tex files.
I gave it my handwritten notes, together with a .tex file I am currently working on and an old project written in the style I want. I asked it to incorporate the handwritten notes, that I shared pictures of, into a lemma while preserving the writing style, and it did pretty well. On a second run, I asked it to refine the proof while following the writing style of a different proof, that was already polished, and it actually understood the assignment!
I mean this is great, can go back to doing math on a chalk board and not worry about the typing.
I used Claude code for latex in vs code for the first time today and it is pretty decent.
While I don't trust it with the math I am doing, I attached pictures of a handwritten proof and asked it to formalize it and write it on my document. It did pretty ok.