What single observation or measurement that has not yet been made (perhaps because it is functionally impossible) would have the biggest impact or provide the most useful data, were it to be made?

Yeah I haven’t watched in ages but still have a lot of their recipes in my rotation. Their gochujang jigae recipe has, as advertised, helped me finish countless tubs of gochujang.

fellow future dish enjoyer 🙌

Certainly this is true!

And I really don’t think my priors on AI, or funding priorities under this admin, are such that I’d be motivated to rationalize this. Which isn’t to say I’m correct!

I don’t totally agree with this type of research receiving this level of funding support, but reading through the order it just doesn’t read to me as a quixotic “computer discover theory of everything” but more in line with the types of things I linked to that I’m already aware of.

Given the reference to ‘foundation models’ I don’t think they mean exclusively LLMs — I imagine the research is going to be more along the lines of stuff like www.anl.gov/article/buil... or arxiv.org/abs/2510.07684

Amazing!

Salesforce tower = Barad-dûr, it works perfectly

thehill.com/homenews/sen...

Statistical mechanics is probably the most bang for your buck if the total number of physics classes is a few. Beyond being very important for astronomy and astrophysics, it can also help provide a foundation for statistical methods (e.g. Gibbs sampling) that may be important to know.

In order of increasing experimentation:

corn bread
fritos
cheeseits

The wheel turns, and it crushes us, but it turns again, and the fight was not worthless merely because it was lost.

Because the cause endures.

EQUALITY OF MAN BEFORE HIS CREATOR

fuck em

thanks! had no idea this was going on