caconym_

At some point during that primary season I was driving in my car listening to my local NPR station and they had a story about these two candidates. I shit you not: they were introduced as "Hillary Clinton" and "the socialist senator from Vermont".

For some reason I remember that moment as vividly as I remember pulling out in the car with my mom on the way to school and hearing that a plane had hit the WTC.

I think Halo could successfully get the Andor treatment (intentional departure from the tone and scope of mainline entries) in the right hands, but so could pretty much anything, and such hands are hard to find.

You generally don't need to do any transcoding if you're only streaming to devices on your local network. I have an old 1060 set up for my Jellyfin server but it pretty much never gets used.

Not OP but I do still call it Facebook.

> the wealthy don't universally aspire to be evil overlords hoarding resources at the apex of a dystopian society

I wouldn't make any such claim (the universality of it, anyway). But in a sci-fi scenario like the one I describe, I doubt "the good ones" are the ones who come out on top. And we can see quite clearly (by their actions if not their words) that many of them do, in fact, want to be hoarding resources at the apex of a dystopian society. "Evil overlords" depends on where you're standing, I guess, but I don't think "dystopian" can be considered controversial anymore.

> to be the first ones up against the wall when the revolution comes?

Well, that's exactly why the possibility of revolution must be managed. See: ubiquitous surveillance, automation and mechanization of the police and military, and general disenfranchisement of everybody and anybody in anything remotely resembling a middle class.

This (implied) end state isn't inconsistent with my scenario. We already see the incredibly hubristic "I'm the only person who can safely handle AI!" attitude among the people working at/running the big labs.

I don't understand this comment. "There will be no real need to produce goods, except as many as the AI owner class will need" is exactly the point---people not in that class will be powerless, useless (to the owning class), and expendable.

The problem I have with the "benevolent billionaires" thing is that it's demonstrably not how they act today.

I think there are plenty of ways for a couple of guys with god-machines to one-up each other, if it comes to that. Feats of technology, longevity, space exploration and colonization, etc. Likely some of these things involve keeping some people around, but billions? I'm not seeing it.

Why do you need to "make money" if you already own a machine that can produce arbitrary goods and perform arbitrary services (including military/security services) for you? You don't need labor for anything anymore. The rational thing to do is to command your machines to dispose of them and reclaim their land and the resources on/under it to feed your machine. Especially since at this point they will want to see your head on a pike.

This is a simplistic scenario that's science fiction for now but I haven't seen anyone make a convincing case for why the underlying logic isn't valid.

> Flock is intended to be searchable by law enforcement, not Joe Average.

Isn't Flock's alleged legality predicated on the principle that there is no "reasonable expectation of privacy" in public? I don't see why that would change depending on whether it's "Joe Average" or a police officer at the controls---IIUC I'm allowed to capture images, video, etc. of people in public just as much as cops are.

Agreed that these systems simply need to be banned.

Trump is not really a dictator due to constraints outside his control, but he certainly behaves as if he wants to be one, and this is the result: it's hard for people to believe that policy moves (or whatever you want to call it) like this are being motivated by a real concern for American workers rather than personal pique or corruption, because not only have so many of his moves been transparently motivated by personal pique and/or corruption, but he's also been restructuring the government in plain view to enable his personal pique and corrupt interests to dictate policy.

The Anthropic thing is the same: we can argue until we're blue in the face about whether the designation as a supply chain risk is legitimate based on Anthropic's policies and actions, but Hegseth is such a vain, insecure clown that any decision like this is suspect.

In short, Trump may not have achieved his dream of dictatorship, but he runs the parts of the government he has control of like a dictator would, and this has done terrible damage to the federal government's legitimacy.

> AI merely exposes the inherent flaws in how we do education (exams, especially failure) [...] but I also wouldn't be able to suggest a framework where students aren't rewarded for skirting around being educated

As someone who graduated from college in the 2000s, I've been surprised to hear from recently-graduated colleagues that doing exams in person is no longer the default. That strikes me as the obvious first line solution for (at least) the STEM subjects that lend themselves to exams centered around technical problem-solving: exams in person, no phones or similarly capable devices, cheaters get a failing grade with no possibility for appeal.

The conventional wisdom when I was in school was that classes where homework isn't graded aren't really any easier for most students, because the homework is where you learn the skills, and if you don't do it then you'll fail the exams. Frankly, the same will be true either way for students who use LLMs for all the exercises: they won't learn the skills, and they will fail the exams. This is a working-as-intended outcome.

So why not start there? It's how school has ~always worked, so there's no question of feasibility. The only question is whether our will to give committed students a real education is stronger than our desire to productize education as a rent extraction scheme tailored to a nepotistic and shallowly credentials-obsessed society where having 'B.Sc.' on your resume sorts you one bucket higher in the algorithmic hiring queue to flip burgers at your local McDonald's.

---

It's less obvious how to fix the humanities subjects where take-home writing assignments carry most of a student's grade, though I'm tempted to think that even a relatively simple and tightly scoped writing exercise given as an in-person exam could conclusively weed out most of the students who've been relying on LLMs to do the assignments for them.

> Didn't those all involve major crimes?

I'm not sure about all, but many did (at least, a serious crime was alleged), which is why I described them as "far worse examples". But they are examples of the same pattern (cops blindly trusting the surveillance oracle), so severity of the alleged infraction/crime apparently won't protect you in either direction.

> Why is the "scrap of paper" a problem now

You misunderstand. I am attempting to present the situation from the point of view of a hypothetical police officer, especially one who may be motivated to view the information in front of him in a certain way that will justify him writing a ticket.

> if the cop thinks it's fake they can call the insurance company and confirm

I expect that the cop in question would not have been willing to do this, if asked to.

> In Georgia, where one of the incidents took place, the cops can lookup insurance status in real-time

I expect that the cop in question had no interest in doing this, or in divulging to the victim that he had the capability to (if, indeed, he was aware of the possibility himself).

Neither of the former two facts is good, but they are facts nonetheless, and holding the officer in question accountable for the error or related broader misconduct will have no preventative effect, because it's impossible to hold American police officers accountable in any meaningful way (i.e. a way that serves as a deterrent in future situations).

---

The first part of the point I'm trying to make is that if we give cops access to surveillance software that purports to tell them whether or not somebody is driving with insurance, they're inevitably going to treat that data as ground truth in many cases. It doesn't matter that you think this officer should have taken the paper as authoritative, or gone to these other sources to verify it: he failed the test and trusted Flock instead, because Flock is a slick commercial product consciously designed to make its users trust it and use it more and rely on it more.

Saying 100% of the blame lies with the cop here (as the comment I originally replied to does) is pretty similar to saying 100% of the harm resulting from teenagers getting addicted to social media is their own fault, and the tech companies who designed their platforms to be addictive to teenagers and chased growth at any cost are blameless. Why don't the kids simply not use social media? It's so easy! Personal responsibility!

The difference is that in the case of cops and Flock (etc.), the harm is turned outwards onto other innocent people.

The thing you need to realize is that reality is the way it is and as a result cops are going to behave the way they do and there's nothing we can do about it in the near to mid term. You appear to think what the cop did here is stupid, and that means you think many cops are stupid, because we've seen far worse examples of "surveillance platform trumps reality according to cop" stories in the news recently. And if you think cops are stupid in this particular way then it would be hard not to agree with me as far as: they can not, in aggregate, be trusted with access to surveillance tools whose outputs are not 100% reliable.

So whose responsibility really is that? Should the governments that purchase surveillance tools, and the companies that sell them, be held accountable for the (predictable) harms they enable? Or should we ignore them, ignore the fact that we all knew this was going to happen and the fact that it keeps fucking happening, and just blame the individual cops? (which, as we all know, amounts to nothing.)

The fabric of our society continues to be eroded by these big moneyed interests because we (the public, the parts of our government that are supposed to protect us from predation) continually and happily swallow all the disingenuous mealy-mouthed bullshit they spew to deflect past, present, and future responsibility for predictable harms. If we keep swallowing then we are headed for bad times. I am done swallowing.

Why should a police officer trust a scrap of paper somebody could easily have faked vs. a massive networked software system that claims to have the up to date ground truth? What if, for instance, the victim got her insurance cards and then stopped making her monthly payments?

(Incidentally, I assume paper insurance cards will cease to be considered valid before too long, and handing your unlocked phone to a cop will be coercively normalized.)

Cops aren't intellectually inclined (or equipped, for that matter) to second guess a surveillance tool that's showing them an unequivocal answer to a question. This is 100% on Flock for selling a tool to users they know won't reliably question its output, despite whatever bullshit CYA terms and conditions come attached to it, and then failing to ensure that its output is correct.

None of the things named in your comment, on either side of the line, really matter.

Reading, writing, and critical thinking skills matter. You don't have to "write structured interpretations of 1950s texts [you] have no emotional connection to" or practice "handwriting, and reading stuffy texts from front to end" to develop those skills (very loaded language, nice job!), but you have to put a lot of time into doing something, ideally somethings, where you're the one doing the heavy intellectual lifting.

Or, we can just write these kids off as a generation incapable of thinking or building by themselves without AI tools that do the thinking and building for them. Honestly, it's probably already too late. The way we've allowed kids of all ages to be exposed to these tools is basically equivalent to handing out cartons of cigarettes in a preschool.

Sounds like a self-fulfilling prophecy. If you draw a line between capital and consumers, and the latter are so tightly squeezed as a result of the "privilege" imbalance that they can barely afford housing, health care, child care, etc., how do you expect them to engage in entrepreneurship? How do you expect them to efficiently allocate their labor?

Meanwhile we see an absurd consolidation of capital that leaves consumers with fewer and fewer choices for basic products and services, allowing capital to make those products and services worse and more extractive. Rinse and repeat. The contempt for consumers and attitude that capital should be "privileged" manifests in our government's total indifference to the former's plight, despite the well known fact that they are the engine of our economy.

How do you think it's going?

I really truly honestly am not sure what to make of this result from $20M in compute, 10K+ parallel agents (smells like brute force), and a pre-existing approach that was already bearing fruit. I know the models are good---I use them every day and continue to be impressed---but how much better than the benchmark of the best publicly available models is this supposed to be? It seems impossible to say.

To be fair, I don't expect anything electronic to survive going through the wash unless it has hardcore waterproofing (eg a dive computer). I was pleasantly surprised by the way my airpods kept working.

I don't remember whether my first airpods were the gen1 pros or the gen2 pros, but you may have been a victim of early teething issues. I don't really have anything negative to say about the ones I've owned.

> overengineered sound makers that breaks the moment you drop them.

Which ones are those? I've owned a couple of generations of airpods and dropped them probably hundreds of times (both in the case and out of it) and it has never had any detectable effect on their appearance or functionality. I even had a pair go through the wash and they still mostly worked afterwards (only issue was that the ANC performance was worse and slightly weird).

Some people claim Astra is significantly better than anything else and significantly more token-efficient, and others (like you) say it's meh and way more expensive to boot. I really don't know what to think.

Kind of a tangent, but one thing I am curious about is to what degree the Navier-Stokes result announced today was primarily a brute-forced result based on the 'program' previously established by researchers to find counterexamples (blowups), or whether the model actually added significant/novel intellectual value beyond its ability to run at arbitrary parallelism. With 10K agents and a staggering $15M in compute (IIRC), I am feeling like a lot of the former may have been involved, but I don't really understand either the problem or the approach (or, indeed, the solution).

Obviously the potential for parallelism and coordination between so many agents is quite scary by itself, but I think brute force by 10K mediocre AI mathematicians is much less scary than ~one AI mathematician reasoning its way through the problem where all human attempts have failed. It seems fairly obvious that massive parallelism lends itself to brute-force counterexample-finding, and I suspect it isn't a coincidence that most of the touted AI math results have been counterexamples.

It's all still quite scary, but coming full circle: I really don't know what to think.

+1. I've used the recent Gemini Flash models and I've used Opus 5, and the latter makes the former look like a box of broken crayons. Unless Flash 3.8 and/or this Muse Spark model are a much bigger deal than people seem to think, I will eat my hat if either one can come close to Opus 5 in actual real life "long-horizon software engineering" tasks.

(I'm not happy about the above being true, but it's the reality I seem to inhabit.)

IMO it will work better in some contexts than in others. If rhythm isn't a concern then yes, it could always be omitted.

it got want to use em dash. but decide: do? q is if appropriate. check martian websner blog. verdict yes---emdash + comma interchangeable---proceed---judgment superficial however no desire dig deeper style irrelevant effect on reader irrelevant meter and rhythm irrelevant restate equivalence with comma established::chain unbroken::consider semicolon? consider ellipsis consider comma consider sentence break all no. preference for emdash est fiat. and all nail shapeds are for hammering.

Well, I disagree! https://news.ycombinator.com/item?id=49527581

It's fine in the sense that when a bad writer writes something I can usually understand what they're trying to say.

Emdashes and commas aren't interchangeable, and your example there demonstrates one great reason why. The emdash establishes a discontinuity rather than one thing flowing into another, which is why the tomatoes don't merit one but the Ferrari does: you are using the emdash to emphasize the situational irony.

Going back to Anthropic's post:

> They’re the world’s most advanced models for coding and knowledge work---and their research capabilities offer an early glimpse of how AI models will contribute to scientific progress.

The first thing directly implies and flows smoothly into the next---or would, if not for the awkward emdash. There is no discontinuity, no twist or shift in context, no implied question and provided answer, no punchline. It's just distracting.

You seem to be fixated on picking apart the wording of my original comment, so let me help you understand it:

> There are obviously few or no hard technical blockers preventing Valve and the Linux community from supporting any game that runs on Windows

'Few' there is doing a fair amount of work: it allows that some significant (or "hard" if you prefer) blockers may still exist (this is consistent with my own observations of game compatibility in aggregate). So your (implicit) insistence that I claimed there are no blockers is obviously nonsensical already. But, furthermore, I went on to say this:

> unless it's been specifically engineered to block Linux

Which precisely describes a game that previously worked on Linux, with anticheat even, but now has been updated to remove that compatibility! It also describes a game whose anticheat has been configured that way since launch.

Also, your implication that "hard technical blocker" and "not worth it to implement" are interchangeable is ridiculous in any scope, whether you're Valve deciding whether to commit tens or hundreds of man-years to Proton development or a one man shop considering a hacky workaround in a client's dated web framework.

You are just barking at the mailman. I'm done here unless you can respond to the actual substance and nuance of my comments rather than a strawman.

Yes, I know what an emdash is---I've been using them in my writing since long before they came to the fore of the AI writing conversation. Anthropic's use of the emdash in the fragment I quoted is clumsy and reads poorly relative to the obvious alternative, a comma.

> We’re introducing Claude Fable 5.1 and Claude Mythos 5.1. They’re the world’s most advanced models for coding and knowledge work—and their research capabilities offer an early glimpse of how AI models will contribute to scientific progress.

I'm not an emdash hater but this isn't how you use them. It should be a comma.