Also when they use cli you can learn from it and sometimes, next time you’ll be faster if its just a simple edit compared to an ai that first has to understand relevant context.
No wonder they think they’re close to AGI when they think we are that stupid.
> The AI System will not be used to independently direct autonomous weapons in any case where law, regulation, or Department policy requires human control, nor will it be used to assume other high-stakes decisions that require approval by a human decisionmaker under the same authorities.
This whole sentence does do absolutely nothing its still do what the law allows you. It’s a full on deceptive sentence.
Altman must have read a lot of Kissinger. If your brain scans the text quickly it almost seems like it's Anthropic's red line, except the second half completely negates it. Completely untrustworthy IMO, this is a direct, malicious intent to misdirect.
Doesn’t matter what they Believe. Not like we are going to do anything about it. Next couple weeks most of HN will be lining up to use the new OpenAI model that’s .01% better.
Perhaps, but there is a difference in a reasoning system deciding on the best way to achieve the goal.
To get the predicted disastrous effects you need to be doing function optimisation without regard to the meaning of the function parameters. Yes, models can still game the system at inference time, but in much the same way as a human might game the system, it requires awareness that you are going against the intent of some rule.
I strongly disagree. It's easy to utter this string of words, but it's meaningless. It's akin to saying if you have two hands you can perform brain surgery. Technically you can, practically you cannot, as there's other things required for pulling that off, not just having two working hands.
I doubt "stopping it" is up to anyone, it's rather a phenomenon and it's quite clear we're all going to wing it. It's a literal fight for power, nobody stops anything of this nature, as any authority that could stop it will choose to accelerate it, just to guarantee its power.
It is not AI we should fear, it's humans controlling and using it. But everyone who has a shot at it is promising they'll use it for "ultimate good" and "world peace" something something, obviously.
Gunpowder (weapons) and atomic tech (energy, material, weapons) are heavily regulated in most of the planet, as the risks of having free access to them for everyone (company/person) for their own selfish purpose without strong guardrails clearly outweighs the benefits.
The fact that something exists doesn't mean that having it readily available is the only option, particularly if it has potentially disastrous consequences at scale. We are choosing to make it available to everyone fully unregulated, and that is a choice that will prove either beneficial or detrimental to society at some point.
I don't think it is inevitable, I think it is a conscious choice made by a few that have their own and only their own interests in mind.
As a technologist, I am amazed at this tech and see some personal benefits. As a human, I am terrified of the potential net negative effects, and I am having trouble reconciling those two feelings.
The challenge is that enforcing a ban would presumably require strict incursions into personal freedoms organized at a scale where AI-based solutions would be particularly effective and thus tempting, paradoxically.
On the other hand, assuming the dangers are real, you lose by default if you do nothing.
One cannot (in most of the planet) go to the supermarket and buy an M16 and a box of hand grenades, or get a hold of a couple of kg of plutonium cause they want some free energy at home.
We also have rules in place of what one individual/company can and cannot do from the point of view of the greater good. I cannot go and kill my neighbour for my benefit (or purposefully destroy his life) without consequences. A myriad of things are not allowed, and I don't see people complaining about any incursion into personal freedoms.
The reason people have accepted these is that we have already proven that having access to those things could be catastrophic. We haven't proven hat yet with AI.
But I don't see much difference between those established and well accepted rules, and a rule that says: A company cannot release or use for its benefit a technology that will impact the need of humans at scale, because of the impact (again at scale) that it would have in society.
In other words, if you are a company and have the potential to release a product, or buy a product from a provider that would cause mass unemployment, should you be legally allowed to do so? I do not think so.
That’s a fair objection. Having ruminated on it some more, I’ll admit it might be tenable.
As for achieving an effective ban, occupational collapse might be the stronger motivator once workplace adoption broadens and accelerates, but risk of epistemic collapse might register sooner among the general public, already broadly suffering slop.
Like Bill Gates, I wonder why it’s not yet become a theme in mainstream politics.
It's doom and gloom because the underlying game theory forces all state actors into an unbound and irresponsible arms race, consequences be damned.
AI development game theory is extremely similar to the game theory behind nuclear arms development, but worse (nuclear weaponry was born from Human General Intelligence, and is therefore a subset of the potential of AI development). Failing to be the most capable actor could put one in a position of permanent loss of autonomy/agency at the whims of more capable actors.
Not OP, but AI is fundamentally in another category than any other technology before it. It requires moral fortitude to wield in a way that guns and books didn't require. It augments human judgement in a way that needs a moral framework to clearly guide it.
Unfortunately, as a species we seem to be abandoning morality as a general principle. Everything is guided by cold hard rationality rather than something greater than us.
The current fruit is automating away a ton of human labor with no foreseeable way to continue to engage that labor. It is poison for the majority of humanity which will bear fruit for the limited few who can use it / own it.
Because it's a fruit governed by humans, in the scope of a capitalistic and patriarchal society. And all fruits planted in a capitalistic and patriarchal society are poison
AI isn't ruthless, that doesn't even make sense. It's a mathematical model, if it's optimizing for the wrong thing then that's strictly the fault of the people who chose what to optimize for
You need to go back and research AI safety long before LLMs were a thing. Any complex goal driven system will have outcomes that cannot be predicted. Saying "it's a mathematical model" belies your ignorance of behavior in complex systems. Very tiny changes in initial conditions can have vastly different outcome in results and you don't have enough entropy in the visible universe to test them all.
I love how sci-fi warned us against hyper-competent galaxy brain conscious AI but we are actually going to be killed by confidently wrong stochastic parrots.
There is no Secretary of War. The name of the Defense Department is set by statute that has not been named regardless of Pete Hegseth's cosplay desires.
Am i the only one who understands the deparments position? Like if another country will have it without safeguards, why would I not want it without safeguards. I can still be the safeguard, but having safeguards enforced by another entity that potentially has to face negative financial consequences seems like a disadvantage, would be weird to accept that as department of war.
Haha I agree that my opinion is kind of that
But more like ports & adapters for semantic space, not just IO boundaries.
If we can abstract the tools one layer further for ai, it might reduce the attention it needs to spend navigating them and leave more context window for actual reasoning
If I'm an insider with 100% confidence, I'll take all offers at a certain price as long as I can afford it. Similar story for lower levels of confidence (but still inside info). There won't necessarily be any left for you to copy at a viable price.
Because there's always some uncertainty and capital limits. But the uncertainty about the outcome is itself inside info, and that's compounded with your own uncertainty about the insider as a copy trader. So the insider will empty out certain price levels only, and your certainty is strictly less than theirs, meaning you have even fewer viable levels to buy.
therefore, the polymarket betting odds will reflect the truth - even if that info is a secret that nobody else but the insider knows. And if this is the case, then even an outsider could make use of the odds as a source of info which would ensure that market efficiency (which is about the flow of information) is high.
If you believe Polymarket as a serious source of truth, consider that somebody manipulated "Will Jesus Christ return before 2027?" because there was a secondary market on whether that market will rise above 5%. Which defeats the whole idea that the betting odds will reflect the truth. Also even pre-manipulation I don't think a 2% chance that Jesus will return was reflective of truth.
The issue comes from situations where the insiders can alter the answer to help their own bets. The simple example is the bet on how long a press conference will be: It's a ridiculous bet when the person giving said press conference can bet and fleece the market.
Will X country invade another before or after day X? A large enough market changes the answer, as the agent can change the decision. And we can see this kind of thing in many interesting questions.
These are not secret divinations though, the participants know this and price it in or otherwise allow it to determine which markets they participate in.
That someone with inside information will e.g. make 500% while those late to the party e.g. only get 10%? (of course your example is not very realistic to begin with)
Has there ever been any documented circumstance where significant inside information became public and known thanks to a trade? Most often, the trade is made at the last minute, and the information gets subsequently revealed anyway. And it's impossible to tell whether somebody is an inside trader, a wealthy gambling addict making a stupid decision, or hypothetically a foreign agent pretending to be an inside trader to make people believe in a particular outcome.
It's impossible to know anything for certain; almost everything is probabilistic.
Also I'm not sure how to interpret your criteria because timing matters, I don't think saying 'it gets revealed in the end' is very meaningful.
Anyway, on Polymarket specifically, sure, military strikes are a common one. Seems like a useful signal to go hide in the basement. Outside Polymarket, there were insider trades in 2008 that I'm sure were useful.
No vigilant insider is making a series of "single market predictions with high accuracy" on the same account. They would make unlinkable bets on fresh accounts.
reply