raf

Not figuring everything out

[raf @rafagnos]

nothing human makes math in the near future

[Bartosz Naskręcki @nasqret]

About pain and grief of mathematicians in the face of new results of OpenAI and others.

In this time of sudden acceleration in mathematics, it's worth reminding all my fellow mathematician friends of the work of Elisabeth Kübler-Ross. There are five stages of grief:

1. Denial
2. Anger
3. Bargaining
4. Depression
5. Acceptance

I've already made it through all of them, and I'm living my happy ever after. I don't think I'm even the same mathematician I used to be and that's okay.

So, to those who are still somewhere between stages 1 and 4: get your best LLMs and learn the new landscape. Learn the new, weird skills of the AI era. Stages 2 and 4, especially, can be very turbulent. Go through them, find new ways of doing the mathematics you love, and move forward. The sooner, the better.

There are only very few of us anyway (probably less than 300k), so let's preserve our human resources. And let's teach the next generation that they will have to discover their own new ways of doing mathematics.

The dinosaurs are gone. This is the era of the swift cockroaches. They survive everything, and build new empires.

[raf @rafagnos]

@letonyo would you consider quantum networks safer?

[raf @rafagnos]

@TeoriaNeutral situational awareness

[Remarks @remarks]

JUST IN: MySpace plans to relaunch the platform

[Noam Brown @polynoamial]

An internal version of Astra, @OpenAI’s next major model family, solved 10 major open problems in mathematics, quantum complexity, and theoretical computer science.

We believe it will be a major step for scientific reasoning. https://t.co/iP6cyheZ7i

[raf @rafagnos]

at least there's more chances of sticking out there...? i guess?

[raf @rafagnos]

gm moots i got rejected by the masters degree i wanted but hey the mid one wants me could be worse

[roon @tszzl]

both of the leading labs have had serious loss of control incidents. there will be serious coping about this from both sides and from /acc bystanders but these are complex emergent loss of control incidents that were detected weeks after the fact

[raf @rafagnos]

@TeoriaNeutral training a model in each religion and/or tradition would be a funny way of reaching multipolar singularity

[raf @rafagnos]

i write this while digging ofc, scoop more and its done

[raf @rafagnos]

note to self: the rabbithole always deepens, don't stay up till 6 am. sleep.

[Sam Altman @sama]

a lot of crazy stories start with too much amphetamines and end with not enough psychedelics

[Elliot Arledge @elliotarledge]

Introducing Netherite!

Minecraft 1.11.2, rewritten from scratch in C and CUDA, bit-verified against the real game.

This is one trained agent playing in it - then the renderer it actually trains through - then 7,200 live worlds stepping in lockstep on one GPU.

[raf @rafagnos]

@ssi what did ilya ship

[raf @rafagnos]

@macusuweru @Oszillate he is missed

[raf @rafagnos]

expected next-word-predictor behaviour

[Rota 🚪���‍♂️ @pli_cachete]

Probably nothing

[raf @rafagnos]

@singularityHSN @scaling01 he could've quoted it but he screenshotted it he's a hero

[raf @rafagnos]

EVERYDAY I WAKE UP AND I HOPE YOU'RE WELL

[raf @rafagnos]

funny to think if i eventually study for years and get a PhD it'll be to be able to prompt at a high level and to understand what the machine god utters

[raf @rafagnos]

this will be a recurring problem

[Rishikesh Gajjala @publishiperishi]

GPT-5.6 Sol has claimed to resolved one of the five major problems in quantum information theory announced by KCIK (which carries a cash prize).

We (@Quantum_Kishor @TobiasHaug_Q ) sat down to verify its claimed proof and surprisingly, we found another paper published the very next day, with NO AI acknowledgment, containing a proof and phrasing similar to what we got when we prompted it ourselves before the result was announced. An interesting coincidence!