Lightnews — Scholar-powered news

Will Lowe @conjugateprior.org · 2h

I think 'staged effect' is underselling it. Performed candour takes careful, persistent effort to pull off and often only works locally.

When I hear her songs it's always like there's some complicated deus ex machina lowering her into the foreground to talk to me. That's ok; it's a cool machine.

1 1

Will Lowe @conjugateprior.org · 1d

Not really the point, but it's a bit weird to say that Bayes can be 'gamed' by choosing priors to support your conclusion. I mean, how is that gaming? Where's the subterfuge? Do we not make fun of villains who elaborate their plans to the viewer?

1

Will Lowe @conjugateprior.org · 1d

fwiw I would agree they're are not _exactly_ the same, but I'd also maintain there are counterfactuals in both and the issue is whether the second set of counterfactuals is sketchier than the first.

There are various high theory answers to that, but tbh I'm more interested in your intuitions.  
3/3

Will Lowe @conjugateprior.org · 1d

or, if that's not doable, explaining how the counterfactuals invoked in it are different from the ones in, say "the difference between the observed average outcome for treated units and the average outcome had none of them been treated (although they were)", to take an ATT example
2/3

1

Will Lowe @conjugateprior.org · 1d

You're very welcome. Articulating what 'another level' means would be useful.

I think the challenge is to spell out, say, "the proportion of observed statistics more extreme than the observed one if the true parameter took the value 0 (which I think it doesn't)" without any counterfactuals  
1/3

1

Will Lowe @conjugateprior.org · 2d

Hmm, gives a new meaning to having "a two-pipe problem"

1

Will Lowe @conjugateprior.org · 2d

"pre-packaged mulled wine *worth* £110k" is either an amazing scientific breakthrough or just the usual stuff, but in such large quantities that they're going to have to use Malham Tarn to store it.

1 1

Will Lowe @conjugateprior.org · 2d

Weirdly the criteria for naturalization now seem to be the same as for permanent residency which, from the immigrant's perspective, marks an interesting values-based choice point

(There's a relevant convo happening in the UK about ILR but naturally neither country is taking any notice of the other)

𝚃𝚊𝚛𝚊𝚜 𝙶𝚛𝚎𝚜𝚌𝚘𝚎 🚇 @taras-grescoe.com · 2d

“We assume that car use is an incompressible liquid that must be routed somewhere. But it’s more more like a gas that fills whatever space it is given.”

—Ian Lockwood, Harvard Loeb Fellow, transport planner.

12 150 530

Will Lowe @conjugateprior.org · 2d

Fair. Electoral gain is an empirical question. Frankly I doubt anyone is trying to measure anything that might tell them, both because that's hard and because pure vibing is traditional on immigration issues.

I would note that none of this affects ease of immigration, just later naturalization.

1 1

Will Lowe @conjugateprior.org · 2d

DIN 476, mfers

Gernot Wagner @gwagner.com · 2d

That's the key, yes:

Brain drain = missing brain gain

Won't make for splashy headlines, but let's just say I'm writing a lot of recommendation letters in A4 instead of Letter-sized format these days.

Thijs Heus @thijsheus.bsky.social · 2d

You won't notice those, is my guess. Those are the international postdocs/PhDs who now choose to go back to Europe/China/... instead of taking positions at the flaghsips.
Much of the US brain drain is going to be a lack of brain gain.

2

Will Lowe @conjugateprior.org · 2d

I'm secretly sort of relieved that this was the sacrifice

"Look, look we are sooo tough, and now watch as we *smite* this bit of legislation [cheer now please] and reap the votes, having slowed down 500 people's applications for two years (which is about how long it takes in Berlin anyway)"

1 1

Will Lowe @conjugateprior.org · 2d

[record scratch]
I expect you're wondering how I found myself putting "Duflo" on my mute words list.

Well... here it is.

4

Will Lowe @conjugateprior.org · 2d

I'm not sure how attractive it could have been given ~500 people seem to have applied for it.

For (afaict mostly hypothetical) US arrivals, the EU blue card and ordinary residency setup seems fine, modulo standard Amt intractability. Allowing multiple passports is the low key game changer imho.

1 1

Gernot Wagner @gwagner.com · 2d

That's the key, yes:

Brain drain = missing brain gain

Won't make for splashy headlines, but let's just say I'm writing a lot of recommendation letters in A4 instead of Letter-sized format these days.

Thijs Heus @thijsheus.bsky.social · 2d

You won't notice those, is my guess. Those are the international postdocs/PhDs who now choose to go back to Europe/China/... instead of taking positions at the flaghsips.
Much of the US brain drain is going to be a lack of brain gain.

1 12

Will Lowe @conjugateprior.org · 2d

Possibly just because I was just rereading transacl.org/ojs/index.ph... but I feel like it wouldn't have hurt them to acknowledge the last 20 years of statistical annotator models. In particular the partially pooled versions.

1 4

Will Lowe @conjugateprior.org · 2d

Introducing... Starter Patches

1

Will Lowe @conjugateprior.org · 2d

And you'll have a nice concrete example of the consequences of differential item functioning, too.

1 3

Will Lowe @conjugateprior.org · 2d

If you're unhappy with structured assumptions about what could have been observed but wasn't, then there's not much left of statistics except counting things. Any confidence interval is populated almost entirely by data that could have been observed but wasn't and is in it (or not) by assumption.

1 1

Will Lowe @conjugateprior.org · 2d

Finally, some descriptive representation.

1 3

Will Lowe @conjugateprior.org · 2d

if you gift it to him, perhaps

1

Will Lowe @conjugateprior.org · 2d

Turns out the 26.1 beta is quite a bit less silly-looking, should you wish to fix the most egregious fails.

3

Will Lowe @conjugateprior.org · 2d

That's certainly something the current Charlemagne knows how to write.

1

Will Lowe @conjugateprior.org · 3d

fwiw my beef with the "writing is thinking" crowd, banging on about LLMs, is not that they're wrong but that they don't take the idea seriously enough.

2

Will Lowe @conjugateprior.org · 3d

which is why this as an identity not a prerequisite relationship: it's the code that realizes your thoughts about a problem's structure and solution. Sometimes those are creative, novel, even inspiring thoughts (sometimes not), but when your brain turns to lunch, those thoughts are still there.

2 2 3

Will Lowe @conjugateprior.org · 3d

writing is thinking (code edition)

1 3 15

Will Lowe @conjugateprior.org · 3d

Wait, ICD-11 is also my mother in a log flume?

4

Will Lowe @conjugateprior.org · 3d

I would also echo the bandcamp suggestion in the other replies. I find great stuff there quite regularly.

And if you want to pay people more serious money for their music, buying stuff there is much more effective than selecting among low (and very marginally different) stream payment rates.

Will Lowe @conjugateprior.org · 3d

Almost every other service had better quality streams when I jumped from Spotify to Qobuz. It was nice and I would have stayed with it had they not incomprehensibly locked the language of sleeve notes, commentary & other text content to that of the billing address. I ended up on Apple.

1 1