asgraham

This has unfortunately been the case in experimental academia forever.

Ehhhh:

1) You can merge a "new change" to fix a merge, instead of reverting

2) It's unclear whether the substance of their individual PRs are the same. Never thought I'd want an LOC metric.

3) Really, I'm going to double down on the idea that merges could be fixing merges. Also happens to increase the merge metric.

1. Yeah, I’m not fully sold on the argument. I’m also oversimplifying it quite a bit by saying it’s just “brainless, repetitive.” That’s just where it’s obvious. But at the same time, my understanding of 95% of the current math results is that they are sort of the result of AI doing dull repetitive theory work, rather than inspired novel thinking. Either way, the article’s point is that the way we use intelligence to do work will change, not that the current cream of the crop will be unaffected.

2. A major and weird point of the article argues that most people are “stupid” because of a lack of incentives for intellectual labor. If you know you’re not that top 1%, why bother in school? If you’ve already missed the window for a fancy white collar job, why bother “thinking”? In this view, AI could have a democratizing effect on intellectual labor.

3. That does seem to be the upshot, but the author doesn’t seem to be very sympathetic to AI execs either (basically calling them rent seekers still in the mercantilist mindset despite developing the “transformative technology”). Still article doesn’t offer any path forward.

But… that’s exactly the substance of this argument. I’m not sold on the full extent of the metaphor and I’m definitely not out here cheerleading AI, but the central argument of this article is this analogy: when you say “but humans work to eat” you’re echoing exactly the concerns during the Industrial Revolution (slash advent of capitalism, depending how you and the article want to look at it).

Will that analogy hold? Who knows. But it’s more than a facile AI bro saying AI is bringing us a utopia. The article even raises a key part of the analogy: the transition to industry and capitalism (and the resulting freed up labor) led to an unprecedented wave of revolutions and death.

That’s not really what it’s saying. The analogy it draws is between subsistence farmers and subsistence thinkers aka white collar workers/paper pushers etc. The argument would be that repetitive, brainless thinking will no longer be required (not thinking in general), so people will be freed up for currently unimaginable cognitive pursuits.

Most people need to be reminded of this. Repeatedly.

I would be shocked if there’s an agreed upon list, certainly not one supported by neuro recordings. You’re probably well-served by coming up with a good set by your own informal observation.

To clarify slightly, K and V are used to construct the dictionary, and then Q is the lookup key D[Q] as described (with the lookup being based on similarity rather than exact hash as it would be in an actual dictionary).

I’m not deeply pro-codex, but it does have all those features. Who knows since when.

My question from (2) would be: is Symbolics still not on par with SymPy?

I don’t know why you’re getting downvoted. If they say “you’re giving us your copyright when you press the POST button” then that would seem to be that.

Seems like a bad place for them to be, given the whole argument of “you can’t sue us for what other people said on our platform, that’s not our speech!” but that seems complicated and IANAL.

Clients are not proxies. The fundamental difference is who accesses the target server. If you’re using a client (in the sense of a local program that renders your request, like a browser or a browser with extensions) then you are the one accessing the information, and once you have that information it’s hard to make a legal argument against transforming it however you want for your own consumption.

But a proxy means you request information from someone else, not from the target server, and the proxy requests from the target server. This is also fine! Except: a) the target server can absolutely say “Proxy A is not allowed to access me.” [1] and b) if the target server has copyright on the material being requested, then the proxy may be violating that copyright by making the data available for your request.

[1] there’s a weird middle ground here where if a server can tell you’re using a client it doesn’t like, then technically it could use its arbitrary denial ability to deny access on that basis, but that doesn’t seem to be legally accepted? IANAL. Or maybe that just doesn’t happen because there’s no way for it to tell what client you’re using. They certainly do try to deny service when they can figure out you’re using an ad-blocker.

I have heard multiple people say almost that exact sentence unironically.

Sure, and usually one of the ways Rust serves us is with safety guarantees.

Which isn’t to say this optimization is a bad idea, just to say it’s sort of a straw man to imply coding in Rust to take advantage of safety guarantees is “serving Rust”

Microwaving takes less attention and less skill, though. And less cleanup. And you don’t need to own a pan. And you don’t need to own spices (steak is a bad example ofc). And you don’t need to think about the side dishes. And you don’t need to agree on what’s for dinner. And I think it’s worth repeating, you don’t need any skill and can’t really screw it up. And you don’t need to worry about it going bad because you were too lazy to cook it. Etc.

Processed food may not be cheaper, but the flavor:effort ratio is waaaaaay higher.

But can I still give all my emails to OpenAI??

(semi-serious question because I’m currently very opposed, but I’m afraid of my future decisions)

How good are these detectors? (Serious question— I’ve never had a reason to use one)

Out of context it does sound like AInese. I think that’s largely because AI has a tendency to use location in very metaphorical terms that don’t quite read, and usually the idea of “relocating” a word would be exactly that kind of vaguely weird metaphorical use of location.

However in the context of this article “relocated the word” is both literal and precise. The word’s target is in a different physical location.

“Quietly” is also AI-coded but does fit in the overall aesthetic of the article.

Can you recommend a specific home assistant community to check out?

To be clarify for everyone: both of the cited articles argue in favor of HPV vaccination. I assumed they would be arguing against since the comment is arguing against, but that is not the case.

In particular, you've mistaken the result of the second study.

> The vaccinated patients were even older, which makes it even worse, because those are older women who have had more time to be pregnant.

^ this is incorrect. Indeed according to the study, the vaccinated patients are younger (33.1) than the unvaccinated patients (37.4), which could easily explain the difference in gravidity. The authors do not report having controlled for age when computing the gravidity effect.

Note also that the entire study was conducted with a population of patients seeking fertility care, so the study can't support the general claim "gravidity is halved for the HPV vaccinated" even were the significance level to survive age-controlling statistics (which it likely would not).

Sure, sure. I’m not arguing for removing drivers for uncommon devices, or even rare devices. But there’s a line somewhere. Maybe it’s at “devices that no longer exist.” But I think it’s somewhere before that. And I have no idea how you’d figure out which devices fall where around this hypothetical line. I can only hope that they had good justification for these removals.

Is the parent really being sarcastic? I read it as genuine.

There’s presumably plenty of code bloat in the kernel, and while no human would ever scan for bugs in a corner of the kernel that hasn’t been used or touched in decades, AI 100% will. And while those bug reports might be useless as bug reports, they seem promising as “why is this code even here?” flags.

> They can determine how much each work contributed based on those weights, so it's dishonest for them to argue it isn't possible.

I don’t know about impossible but it’s definitely not a straightforward read from the post-training weights as you’re implying, unless you’re aware of some technique I’m not aware of.

The closest you could get would be the weight differential from training with a given work. But that’s massively dependent on training order, so that it’s certainly not at all a good measure of “contribution.”

My first impression when the Leaf image loaded was that you were being overdramatic. The Ferrari website created the impression of a similar but fundamentally more elegant car (not elegant, just more elegant).

Then the Ferrari image loaded. Wow.

It really is a game of spot the difference. A difficult game.

edit: I don't want to reduce hypercars purely to their "Wow!" factor, but a huge huge part of their value is definitely the feeling they evoke when you see one out of the corner of your eye and your head snaps around. This Leaf/Luce side-profile similarity is completely antithetical to that "Wow!" factor.

Ignoring the spiritual part, emotional state does have a well-known feedback loop with physical state. There’s a (largely incorrect) idea in pop psychology that just as happiness leads to smiling, smiling leads to happiness. It’s not nearly that simple, but there are some more straightforward examples: lots of tense emotional states (anger, anxiety) lead to tense muscles (jaw being the classic example). Relaxing your jaw can lead to a (temporary) relaxation of your emotional tenseness. I’ve never heard of a similar result for the lower back, but it’s not hard to imagine. If nothing else, they must be correlated through sedentary lifestyle.

Requires the MacOS desktop app. Windows "coming soon," but still no word on an official Linux app.

The good is hidden: court systems are already overwhelmed. If the arbitration cases were added, then it’d take even longer to get a court date.

(Which isn’t to say I think the system as it is is good, just that there is a good)

Are you arguing that eventually a competitor will emerge that does support OpenClaw with a subscription model? Wouldn’t that just be more expensive for the exact same reason Anthropic is banning it?

I know for a fact [1] that the neuroscientific discoveries were not independent of physics: the people doing the developing were largely former physicists. They likely didn't cite anything because why would you cite phase transitions or criticality? You learn about them in class as a physicist. I strongly suspect the ecology results weren't independent either, but all the theoretical ecologists I know are relatively young (if mostly former physicists) so no first person accounts.

The part of this that could totally be true is that a clinical application somewhere along the way "independently" "reinvented" it. There's a hilarious collection of peer-reviewed journal articles out there inventing a "new" method of calculating the sizes of shapes and areas under the curve. The method involves adding up really small rectangles. (I think a top comment already mentioned the Tai article [2])

[1] source: my doctoral advisor was a really really old theoretical neuroscientist who trained as an electrical engineer and mathematician. If you want a more concrete example, the work of Bard Ermentrout on neural criticality starting in the 70's or 80's. He read a lot of physics textbooks.

[2] https://science.slashdot.org/story/10/12/06/0416250/medical-...

The irony is that youth are simulatenously the biggest consumers of (new) social media, and the staunchest haters [EDIT: this is directly contradicted by the research article I found below…]. I can’t find the source so take it with a grain of salt, but I’ve read that something like 80% of TikTok users under some age think they’d be happier if it didn’t exist and/or wish it didn’t exist.

I don’t think this is really an issue of censorship to a lot of people (though that may be how it shakes out in the government) but rather of control over their digital environment and sanity.

EDIT: I don’t think this is what I’m remembering, but it has concrete numbers somewhat lower than I thought (48% of teens think social media harms people their age, but only 14% think it harms them personally) https://www.pewresearch.org/internet/2025/04/22/teens-social...