Lightnews — Scholar-powered news

boarders.bsky.social @boarders.bsky.social · 12h

I don’t think Kuhn had that in mind if you read SoSR, but regardless I think falsificationism faces irredeemable problems both as a descriptive and as a normative theory, the most straightforward one being the Duhem—Quine problem

3

boarders.bsky.social @boarders.bsky.social · 13h

Interesting! All I ever hear is naive falsificationism/naive belief in ‘method’. Most scientists don’t seem to appreciate Kuhn as it suggests science is an all too human practice

1 3

boarders.bsky.social @boarders.bsky.social · 1d

I wonder who those 5% of neutrals are

1

boarders.bsky.social @boarders.bsky.social · 1d

Personally I only let my lack of imagination be mistaken for an insight into necessity when a guy from the 70s did it first

1

boarders.bsky.social @boarders.bsky.social · 1d

thank you for your service 🐥

1

boarders.bsky.social @boarders.bsky.social · 1d

Rationalizing the process of who gets academic funding based on anti-intellectualism in order to only give it to what is tangibly beneficial to the public would be as bad an idea

1

boarders.bsky.social @boarders.bsky.social · 1d

I think wanting to make immigration pathways tied to legible “high skills” is foolish and giving in to xenophobia. Many of the immigrants that go on to do incredible things do not arrive already fully-formed. Legibility like this in economic policy doesn’t and can’t capture intangibles

1 1

boarders.bsky.social @boarders.bsky.social · 1d

1 in 5 high schoolers have had an AI family (played The Sims)

boarders.bsky.social @boarders.bsky.social · 2d

Regime accountant weighs in on handling missing $

2

Reposted

Bill McKibben @billmckibben.bsky.social · 2d

America is in crisis.

So it's a small relief that the rest of the world is actually getting remarkable stuff done with energy. We just passed a huge milestone globally!

Now back to work on America.
billmckibben.substack.com/p/something-...

Something extraordinary just happened

But of course there's a but. Or two.

billmckibben.substack.com

10 72 270

boarders.bsky.social @boarders.bsky.social · 2d

This means you eventually burn through your pain receptors on worrying about anticipating questions because you never can and so you settle into the zen enlightenment of letting it unfold and going with where it goes

3

boarders.bsky.social @boarders.bsky.social · 2d

I have only taught mathematics (including mathematics services courses to other depts) and I have adopted teaching inspired by the inquiry method where the class is driven entirely by students answering questions and elaborating on their suggestions

1 1

boarders.bsky.social @boarders.bsky.social · 2d

I’m not sure what you mean about storing numbers up to integer accuracy, but I think it is somewhat clear what it means to have a good floating point algorithm for sine

1 1

boarders.bsky.social @boarders.bsky.social · 2d

Floating point is way worse than one naively expects

4

boarders.bsky.social @boarders.bsky.social · 2d

In the same way, one finds that algorithms for the derivative are complicated because using a naive Euler approach can give less accurate answer using a smaller interval e.g. f(x+h) for large x and small h is often no good

1 3

boarders.bsky.social @boarders.bsky.social · 2d

By the whole point is finding an algorithm that is uniformly good across how we represent floating point so data storage is intertwined with the algorithm

2 5

boarders.bsky.social @boarders.bsky.social · 3d

Let’s not mourn that we no longer have notice period posting, let’s celebrate that we ever had it

1

boarders.bsky.social @boarders.bsky.social · 3d

Obviously the same thing occurs at an elementary level in classical mechanics when we ask if there is an independent way to measure a force or if Newton’s second law is a tautology

1 1

boarders.bsky.social @boarders.bsky.social · 3d

This is one of the points in Hasok Chang’s amazing book Inventing Temperature (i.e. without a theory of what temperature *is* how could one know a thermometer were more or less accurate)

1 3

boarders.bsky.social @boarders.bsky.social · 4d

The way to administer a welfare state also just cannot be flat payments with sharp falloffs

1 2 10

boarders.bsky.social @boarders.bsky.social · 4d

When the room starts writing good stuff, I’ll start reading it

5

boarders.bsky.social @boarders.bsky.social · 6d

It is plausible to me that one could teach via simplistic models for helping students inquires towards understanding for themselves about adaption (as a book on evolutionary game theory might do) - I wonder if there are books that take that approach?

1

boarders.bsky.social @boarders.bsky.social · 6d

Why does everyone sans Grace think that they are trivially smarter than Alan Turing and, more broadly, _any-philosopher with functionalist
inclinations? It seems simply because it fits most conveniently with their other commitments regarding the political economy of LLMs

1

boarders.bsky.social @boarders.bsky.social · 6d

I’m not sure if it is social media or was ever the case, but the casual trotting out of supposed scientific facts which clearly flatter the prior convictions of the person stating them is really unnerving to me

1 1

boarders.bsky.social @boarders.bsky.social · 8d

Longing for the nice and tight labour market

3