ar813

Unfortunately it is entirely related. The initial wave of cancelations were about those topics, but the more structural thing they are attempting now is to focus funds on tool development for AI models as opposed to human capital development as NSF does.

That's wild! Now I want to go hunting for more such examples..

If I take a step back and think back to say a few (or 5) years ago, what LLMs can do is amazing. One has to acknowledge that (or at least, I do). But as a scientist it's been rather interesting to probe the jagged edge and unreliability, including using deep research tools, on any topic I know well.

If I read through the reports and summaries it generates, it seems at first glance correct - the jargon is used correctly, and physical phenomena referred to mostly accurately. But very quickly I realize that, even with the deep research features and citations, it's making a bunch of incorrect inferences that likely arise from certain concepts (words, really) co-occurring in documents but are actually physically not causally linked or otherwise fundamentally connected. In addition to some strange leading sentences and arguments made, this often ends up creating entirely inappropriate topic headings/ sections connecting things that really shouldn't be together.

One small example of course, but this type of error (usually multiple errors) shows up in both Gemini and OpenAI models, and even with some very specific prompts and multiple turns. And keeps happening for topics in the fields I work in in the physical sciences and engineering. I'm not sure one could RL hard enough to correct this sort of thing (and it is not likely worth the time and money), but perhaps my imagination is limited.

If they can do this to Harvard, what hope do other universities have?

Between NIH grant-making being slowed down and this, there is no way to interpret these moves than being an underhanded way of causing university research infrastructure to collapse. Consider also, for example, that one Friday last month there was an announcement that indirect rates being cut to 15% for Dept. of Energy grants.. but only for universities, not companies or national labs!

Just from a steady state picture there is now significantly less funding flowing to America's research institutions and institutes. At some point buildings will be shut down, infrastructure mothballed and a generation of scientists will simply not be trained.

An adversary could not ask for a better outcome.

The answer to this problem lies here: “Entsminger pointed out that roughly 80% of the river’s flow is used for agriculture, and most of that for thirsty crops like alfalfa, which is mainly grown for cattle, both in the U.S. and overseas.”

The simple solution would be to raise prices on water such that it disincentivizes growing water hungry crops than alfalfa for example. The west’s water crisis is less about cities than agricultural choices made during the last century, which was wetter than it will be going forward. The obvious answer is to either regulate or incentivize using less water hungry crops more strongly. It would be better if this had started slowly a while ago, allowing the market to adjust and reallocate. Alas, looks like it will have to be an abrupt shift in the near future.

Erdős apparently came up with the following poem to break the ice, which is pretty funny!

This is the city of Madras / The home of the curry and the dal Where Iyers speak only to Iyengars/ And Iyengars speak only to God.

Was that last line a reference to Ramanujan?

I think the points raised are reasonable and, to my mind, not particularly partisan (though I would argue one party is more responsible for the current chaos by several orders of magnitude than the other one..).

The deeper issue this raises is, I suppose, a constitutional one paired with Long Now Foundation-type questions. How do we design institutions that can't themselves be changed too easily due to majoritarian whims, yet can evolve over decades and centuries as societies change. We certainly shouldn't expect the institutions and norms of today to precisely meet the needs of our descendants decades or centuries from now.

For the here and now though, my observation is this: Americans in general are allergic to learning from what other countries do. But really, as a very large country, we should be studying what smaller countries do and using them as experimental points of data and testing them out here. There are plenty of implementations of better, more responsive governance out there. Nothing's perfect, but saying 'we're number one' and plugging our ears is no solution.

Singapore is similar, in that pay for top public servants is competitive with top private sector pay. And they try to attract the best and brightest.

I don't think it will reduce the need for energy storage. If it does, it might be in some marginal cases and by a small amount. This is because the power generated using this approach is quite a bit less than what you can get from solar. So for any conventional uses, PV+storage will always be the winner. I think the real advantage for this might be for low-power, long-duration applications where battery cycles can be a challenge. The other big scenario is polar climates, where there's low solar insolation for several months.

Yes that distance matters in so far as getting the heat to the thermoelectric can be a bit more challenging. However I believe this has been investigated before and there are likely ways of doing it at least somewhat well.