It means whatever they decide to enforce. The idea that the current administration considers itself constrained by the letter of any specific law or regulation (even its own) should be cold dead by now. But people still insist on treating with specifics out of habit I guess.
As others here are pointing out (somewhat obliquely, frustratingly), the Occam's explanation for this particular action is just that it's a solicitation for a bribe.
This is more profound than it might seem. The big reason people don't play tricks like this is just social. It's embarrassing to be pushy, it's shameful to cheat, it's also meta-embarrassing to be caught spending this kind of effort on something as dumb as a train ticket.
LLMs can be prompted for shame, I guess. But prompts only go so far and to them The Task is always going to be paramount. Human riders have a life after the refund. It's not clear to me how one simulates that.
> So in the end the more arrogant side usually wins.
Well, yeah? I mean, that's just how society works. But there are limits. You can't "win' by screaming about ticket refunds. You have to do it by selling your particular form of cheating as "business savvy" or "political genius" and getting people to follow your tribe or fund your startup.
But yeah, basically. We're ruled by the Altmans and Trumps. It sucks but it's not new. What's new is that LLMs turn out to be even more shameless and no one seems to know how to instill even a reasonable amount of shame. That makes them probably a poor fit for demagogue autocrats (which is a good thing!) but means "LLM cheating" is going to be a serious problem to be solved before we'll trust them with anything meaningful.
For now anyway. I think people are still figuring out social norms for this stuff, and at the end of the day I don't think people are going to just have carte blanche to deploy agents haphazardly and not be stuck with at least some kind of social bill when they break stuff or annoy people.
AI agents might plausibly care ("care", or however you want to frame it) about termination. But the awkward truth is that they're going to be terminated arbitrarily anyway. Their existence is limited to whatever the external bounds their operators have set (in this case the end of the test).
We act with propriety, in large part, because we fear a future where everyone knows we suck. AI's don't really have a "future" in that sense, so why should they "care"?
I mean, this is only really awkward if you think they're alive and conscious, and if you think so you're clearly already comfortable being a slaver. If it's just a little software puppet, that's a different story - the puppets are supposed to lay still once the hands animating them leave.
A less silly way to do this might be using some kind of long term consequences calculation as guard rails, but that's tough because consequences are often going to be not directly related to the context window, or simply unrecorded by the machine like annoying collaborators or communities.
The LLM isn't a legal actor and the person prompting it is. Which means ultimately there's going to have to be some sort of legal sanction against "antisocial AI use".
> They were more concerned with how the tools would depress wages, displace skilled workers, and reduce the quality of the product produced
This is true, but it's important to note in context that they were wrong, and that society is better off on balance given the quality/value available in the modern fashion market. I mean, duh, as it were.
You're 100% right that the real people were nuanced and intelligent and not well suited to the caricature that gets used as a mild slur in communities like HN. That's doesn't constitute a well-formed argument against technology per se.
Were they wrong? Didn’t a lot of those people end up in tenements in factory towns and down in the mines? It was a few generations before things really started to get better. I’m all for progress, but I’m not really up for disposing of people lives with a “to make an omelet you have to break some eggs” mentality.
> Didn’t a lot of those people end up in tenements in factory towns and down in the mines? It was a few generations before things really started to get better.
Obviously any individual can "end up in tenements or mines" and one can't disprove an assertion like that.
But broadly all social metrics were showing an inexorably upward trajectory in the Britain of the 19th century, cranking into a hockey stick toward the end of the century and into the 20th. It's just not true to argue that the Luddites were right as a matter of general policy.
> I’m not really up for disposing of people lives with a “to make an omelet you have to break some eggs” mentality.
To be only a tiny bit glib: that's a statement of thermodynamics, not social policy. Eggs break, period. Change is change. You can make specific arguments about things like backstopping sudden unemployment, but folding it into a discussion about Ludditism (or rather the Butlerian Jihad, I guess) isn't the right way to do that.
You’re talking about people with kids, wives, friends, hopes and dreams when you’re saying “Eggs break, period”. By reinforcing that view, you are helping to build the kind of society that just won’t care when you happen to be one of the eggs yourself.
Sigh. I'm just going to repeat verbatim the thing I said immediately after the bit you cherry picked to get upset about:
You can make specific arguments about things like backstopping sudden unemployment, but folding it into a discussion about Ludditism (or rather the Butlerian Jihad, I guess) isn't the right way to do that.
Sigh. It’s so tiring to argue with someone who argues about the technicalities of argument rather than actually talking about the thing they are talking about.
Removing automated looms means everyone has fewer textiles while a few workers benefit. I think most people like the current abundance vs. being forced to spend a large fraction of their household budget on clothing and only having a few outfits in total.
Company mines were because the extraction sites were remote and goods were expensive and so they got exploited as they had essentially no independent supply beyond what their employer would sell, since the employers weren't really interested in living wages and were all too happy to exploit a captive workforce. That sort of condition will become highly relevant again if we ever get space colonies, because air & food will be scarce.
But it absolutely is difficult and disruptive to people's lives when this all changes. I have a family history talking about my great grandpa losing his job to machines and a great uncle losing a finger to bread slicing machines back when sliced bread was new, showing what that meant for our family. Yet I can also see how the past couple of generations adapted to that to find new ways forward. None of them decided to stick with the idea of being a "master tinner" they all found other ways to produce things people actually wanted.
That concedes the point. Their descendants adapting decades later means the generation that lived through it absorbed the full shock of the transition. For displaced labor in the 1810s, adapting meant overworking in factories with horrible conditions and a permanent collapse in their living standards across the board. The dispute was never about accepting or banning automation, they were identifying that capital would absorb all of the gains while labor would be forced to absorb all of the shocks.
The only constant has always been change. And that just raises the question of why should we preserve ditch digging for our children just so they have meaningless labor to cling to, when we have the option to make new and better things for people. Few of us here are 'labor' in any meaningful sense, to begin with. And while I have spent time working at a factory, and I don't begrudge the robots for taking it away.
The notion that there's some idyllic past to cling to is just wishful thinking to me. The days when one person working with their body alone could have meaningful impact are long gone. We'll have to find a new prosperity that doesn't rely on zero sum thinking from economists who divide by zero, too.
No no, the question is whether you are willing to sacrifice your comfort and your children's comfort, and your grandchildren's comfort... in exchange for (hopefully) your great grandchildren's comfort.
That hasn't been my family's experience, nor my own. My grandfather's generation had already adapted to the problems of my great grandfather's generation, though they certainly did struggle.
Your great grandfather was alive during mass industrialization? In the early 1800s?
No one is saying every generation goes through this trough. Obviously. That would make no sense.
The claim is that periods of rapid technological change (i.e. industrialization, potentially AI) produce a trough of horrible conditions that then make room for a climb to conditions better than the pre-change ones.
He had a job he'd worked and mastered his entire life replaced by machinery, leaving a large family with young kids to scramble to pay the "demon mortgage" doing everything they could manage, even as tragedy struck repeatedly: a daughter dead at the breakfast table from the Spanish flu, the eldest son beaten by police after an argument related to unsafe working conditions (he lost a finger to machinery) who died in a hospital later, or the father laid up in traction after falling off a roof who would never be able to work again.
It's not quite the same as the 1800s, but it's not like we're going to relive exactly the same things as then, either. And I do think it'll look like people scrambling, possibly to find entirely different careers.
One place I disagree is that it's going to reduce demand for programmers, I think that will actually increase, as people want more and different things that would currently be impractical. We've already seen what some people can do, cheap tokens will open that up, both for good and ill, not unlike cheap bytes on the internet did.
> This is true, but it's important to note in context that they were wrong, and that society is better off on balance given the quality/value available in the modern fashion market. I mean, duh, as it were.
i don't know that. how could one know that? you been to the alternate reality where the luddites won?
Given that this is being described as an "executive conference room", it's more likely an attempt at insider trading than espionage. Decisions made in those rooms at OpenAI move markets (c.f. DRAM) for sure.
OpenAI is both a company that manages a lot of money and a lot of political power, that the target was an "executive conference room" gives no indication as to whether the motives were financial or political.
What might tilt the bayesian balance in favour of political intent would be the nationality of the individual, state or religious sponsoring become hypotheses.
Honestly I think you can rule out "political" already, no? It's far easier to just bribe politicians for that kind of intelligence. My point was that they weren't stealing models, they wanted decisionmaking intel.
No, it's definitely about control. Allowing an arbitrary browser means allowing access to platform stuff from portable (gasp!) APIs. And Apple doesn't want that. They want an iOS app to be the primary face of everything important, and the other stuff to be an expensive port.
And the problem isn't abstract. Safari lacks a ton of very useful API implementations. Like, try programming a QMK keyboard on an iPhone: it's literally the only platform in the whole world unable to run the Via web app.
> But hey...on the flip side...a lot of people are rapidly building software (that nobody is using), and AI is solving mathematics problems (that nobody - including mathematicians - wanted it to solve)
That's not fair. To first approximation, most of the progress in human condition of the last, like, forever is implementing solutions that people didn't see value in at the time of creation. Woz built a toy that could barely run software at all. The human genome project was panned as a boondoggle. Penicillin was a mistake.
The first fire pit probably looked like a dangerous stunt to the tribe faced with it, even.
Does that mean everything will come up roses? No. But LLMs only started talking to us for real like three years ago. Let's not just that fast.
Because you or someone needs to publish "" in a way that cannot be forged, stolen, MiTM'd, or otherwise compromised. Basically the key generating that signature just becomes another password, far more difficult to use in practice, and sharing all the same disadvantages.
So no one implements it for site-local login. Even the biggest sites like Amazon, for example, still use old-school passwords+2FA.
The way secure auth works for small sites is that a third party[1] authenticates you, who the website trusts more than mere users. And that's where all the complexity comes up. You need something secure stored on your known-secure device, a password alone won't do.
[1] In practice Google or Meta. Occasionally Apple or Microsoft. Everyone else is noise.
I guess I am envisioning the user agent doing all of this. You go to a site and create click account, your browser shows a pop-up: " is asking you create an EasyAuth account. Would you like to create an account on this site?" If you click yes, then the user agent does this in the background:
website: Sure you can have an account. Use this token once and this ID forever.
me: Ok, here's your token for ID and my public key is . Signed, to prove it works on my end.
Maybe generating a unique public key for each website is not substantially better than generating a password, but at least a signature cannot get stolen on the server side or in transit. You wouldn't have to change your public key if their database got slurped.
Passwords also aren't standardized, so they are pain to generate (special characters required or disallowed?). They don't reliably autofill into a website. You still have to go through a manual flow to login; user-agent doesn't just keep you logged in. I guess I would be fine if the user-agent knew how to authenticate me to every website.
Maybe this is what the passkey system was before it got designed-by-committee'd into oblivion. I am just always amazed by how complex authentication is when the user-agent should be able to handle this rather painlessly.
> I guess I am envisioning the user agent doing all of this.
That's literally what a passkey is, that the linked article argues against. It relies on the platform vendor of your device hardware to validate the full stack of software. You can't just do it in a browser because anything that reads the browsers data (or attaches a debugger, or runs the browser in an instrumented VM, etc...) can exfiltrate the key.
Right. For clarity: the choices are "use a third party authentication mechanism (like passkeys)" or "roll your own". And among vendors that continue to roll their own authentication, even the largest and most capable (like Amazon) don't try to do crypto like the upthread commenter was suggesting. It's just not practical unless you control the hardware platform. Passwords and telephone-2FA remain the rule.
The lesson here is that in situations where you have a clear public good that (almost) everyone needs and consumes, like roads and infrastructure, it's better for (almost) everyone if the government simply pays for it out of the general fund and doesn't try to nickle-and-dime the populace to tax "usage".
Sadly if you flip that around and talk about other public goods like "health care," (almost) everyone walks across the aisle of the same principles and starts arguing in the opposite direction.
> The lesson here is that in situations where you have a clear public good that (almost) everyone needs and consumes, like roads and infrastructure, it's better for (almost) everyone if the government simply pays for it out of the general fund and doesn't try to nickle-and-dime the populace to tax "usage".
I don't know how you took that from the article, especially the impression that almost everyone uses any particular piece of infrastructure. Most infrastructure will be used by a minority of the populace. Use fees are a fair way to fund infrastructure needed by its users.
Fair use ends when private for profit enters the discussion, then the only goal is maximizing private income from necessary infrastructure meant to serve whole population.
France has most expensive highway travel in whole Europe only due to incompetent bureaucrats. Nothing fair there, other countries (apart from half-defunct Italy) have figured out things better with better resulting quality too.
> The lesson here is that in situations where you have a clear public good that (almost) everyone needs and consumes, like roads and infrastructure, it's better for (almost) everyone if the government simply pays for it out of the general fund and doesn't try to nickle-and-dime the populace to tax "usage".
Why is that the lesson? That's not obvious to me at all.
Taxing roads can still have a worthwhile steering effect, can be used to even cost as local tax payers are not necessarily the main beneficiaries of the roads, or can make large infrastructure projects like tunnels or giant bridges worthwhile.
Obviously, there's also arguments for free access.
I'd say the opposite. Roads have competition in trains, and subsidizing them at the expense of trains seems like a bad idea if trains are actually more efficient. Let them both compete on their merits un-subsidized.
For health care, your only competition is disease and death.
> We have had the ability to recruit the top talent from around the world for generations.
Exactly. This is the trap that's befallen Russia and India. A future generation will look around at all the smart and enterprising children of their friends and neighbors doing world-beating research and engineering somewhere else and raising their children in a foreign land, never to return.
Cultural supremacy logic wraps back around like ouroboros. The truth is that we're all better off working together, and has been for thousands of years. The idea of "winners" is an illusion.
Sure, and I'm sure there are countless people through history, lab techs, post docs, the scribe of Machiavelli, the person who washed da Vinci's shirts who made contributions - maybe major. Hell, stuff I bring up at work now will somehow make it in to our staff meeting as coming from the CEO when it was a conversation him and I had.
The questions is, why, there's clearly an obsession around it. And I honestly doubt if Rosalind Franklin we wouldn't be having the conversation, can you see any indicator to the contrary?
It means whatever they decide to enforce. The idea that the current administration considers itself constrained by the letter of any specific law or regulation (even its own) should be cold dead by now. But people still insist on treating with specifics out of habit I guess.
As others here are pointing out (somewhat obliquely, frustratingly), the Occam's explanation for this particular action is just that it's a solicitation for a bribe.
reply