Hacker Newsnew | past | comments | ask | show | jobs | submit | amai's commentslogin

Dirac: "Yes, that is so."

"There is no God and Dirac is his Prophet".

-- Wolfgang Pauli.


Both were pretty neurodivergent. Dirac had Autism and Feynman probably Autism + ADHD .

The idea to use python code instead of other kinds of tool calls is taken from smolagents: https://huggingface.co/blog/smolagents

It is based on this paper https://huggingface.co/papers/2402.01030 and calls this idea CodeAct. The paper is actually from Apple: https://machinelearning.apple.com/research/codeact

So Astra and Fable seem to take this idea to the extreme causing some unwanted side-effects.


I'm not sure the idea is really from a single set place or lineage like that.

If it was, it was at least from before smolagents and those papers - ChatGPT had already been using automatic Python scripting+evaluation calls and people calling it in agentic loops in 2023. The ReAct paper for agentic loops and PAL paper for dynamically calling Python for tasks which can be better done computationally were both from 2022 (but that doesn't mean the idea necessarily sprung from those either, they're just earlier papers published on the topics).


If Trump wouldn't threaten Germany (probably with tariffs), the AfD would already be forbidden. Don't take my word for it. Politicians from the AfD itself admit that:

https://www.welt.de/politik/deutschland/article696e612da4073...


Nitpick: his name is Sebastien

I thought finite time blow-up for Euler equations had already been proved:

https://www.quantamagazine.org/computer-helps-prove-long-sou...


Why do we scroll and not swipe? Scrolling is not a god-given mechanism to read web pages. Swiping/page-turning would be easier on the eyes (text lines don't move) and would probably be less addictive.

In classical Newtonian physics, singularities exist, but only in frictionless systems. One example is the driven oscillator. Another more complex one the https://en.wikipedia.org/wiki/Painlev%C3%A9_conjecture . However, as soon as friction is taken into account, the singularities (e.g., infinitely high amplitudes in the oscillator) are damped.

Fluids without internal friction or viscosity are described by the Euler equations. Therefore, singularities (finite time blowups) occur there, which has recently been shown with a computer assisted proof:

https://www.quantamagazine.org/computer-helps-prove-long-sou...

The Navier-Stokes equations are the Euler equations plus friction/viscosity: As a physicist, I do not expect singularities there, since energy is always lost due to friction.


I don’t think this problem is very relevant. The Millennium Problem only asks about the smoothness of solutions to the incompressible Navier-Stokes equations. Real fluids are compressible and also conduct heat. In other words, if you solve the problem, you only learn how fluids behave under unrealistic assumptions. Physically speaking, one learns nothing from such mathematical exercises. Furthermore, singularities in the velocity field essentially imply velocities greater than the speed of light. That, too, is physically nonsensical, as we know from the theory of relativity. In other words, the solution only tells us that the incompressible Navier-Stokes equations are not realistic. But physicists have known that for a long time.

That is bad. How can they collect important data if the rocket doesn't explode at launch? Space-X does it better.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: