Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Licensing is neither feasible nor effective for addressing AI risks

aisnakeoil.substack.com

Post reply on HN
Print view View original

151–160 of 161 posts

Re: Licensing is neither feasible nor effective for addressing AI risks

#151

Post by anlaw »

drvdevd wrote: post #99
anlaw wrote: post #98

Oh but it is; they can demand hardware based filters and restrictions. “Hardware will filter these vectors from models and block them from frame buffer, audio, etc, or require opt in royalty payments to view them.”

This is an interesting idea but is it feasible? What does “filter these vectors” mean? In the context of deep models are we talking about embedding specific models, weights, parameters, etc at some point in memory with the hardware? Are we taking about filtering input generally and globally (on a general purpose system)?

[dead]

Re: Licensing is neither feasible nor effective for addressing AI risks

#152

Post by mrtranscendence »

pixl97 wrote: post #120
williamtrask wrote: post #115

Earlier quoted context omitted.

> there are no physical or info-theoretical limits that superhuman intelligence must require a megawatt-sized compute cluster and all the data on the internet (which obviously no human could ever see). Much of your "intelligence" is a function of natural selection. This is billions of years X bajillions of creatures in parallel, each processing tons of data at a crazy fast sampling rate in an insanely large/expensive…

LLMs we're messing with are text data only, we're barely starting to eat video data in multi modal LLMs. This world doesn't lack data.

I'm not sure why watching Linus Tech Tips and makeup tutorials is going to give AI a better shot at super-intelligence, but sure?

Re: Licensing is neither feasible nor effective for addressing AI risks

#153

Post by dontupvoteme »

gmerc wrote: post #150
dontupvoteme wrote: post #94

Earlier quoted context omitted.

What about all those GPUs they've already made?

That’s not how growth works. The moment the current subsidized growth stops the ecosystem would come to a screeching halt.

Growth is not always so simple. The countless GPUs previously used for PoW crypto - Where have they gone?

Re: Licensing is neither feasible nor effective for addressing AI risks

#154

Post by DirkH »

throwaway9274 wrote: post #147
DirkH wrote: post #144

Earlier quoted context omitted.

There was also only "a lot of theory" and zero empirical evidence that nuclear bombs were possible not that long ago. Theoretical fact-based peer reviewed literature often (though not always) precedes empirical evidence. We shouldn't jump the gun and say runaway AI is inevitable just because there is fact-based theory on its possibility. Just like it would have been dumb to just look at E=mc^2 and conclude nuclear we…

Prior to the development of nuclear weapons, there was well-grounded peer reviewed theory in the physics literature. There were predictions that could be made by the theoretical frameworks, and experiments to confirm those predictions. The literature well described the composition of atoms, the decay of radioactive elements, the concept of fission, and the conversion of mass to energy. No such groundwork exists to su…

Don't we likewise have predictions about FLOPs, compute, algorithms etc etc that are grounded in theoretical frameworks, and every year or so we make predictions and make experiments to confirm or deny AI capability predictions?

The AI and neuroscience literature likewise describe the composition of intelligence and what it takes to build it.

What would you need to see to believe we have enough evidence that runaway AI is possible, even if it is highly unlikely any time soon? (Worth adding here I don't know exactly what you mean by runaway AI other than simply AI humanity as a whole loses control over). I'm asking what it would take for you to go from 0% possible to 5% possible.

Last time I checked science fiction authors in the atomic age making fanciful speculation on the properties of imaginary systems weren't responsible for actually building said systems. Though naturally, there will be exceptions. Nonetheless, I still have a hard time dismissing Geoffrey Hinton's concerns as fanciful sci-fi.

Likewise, a nuclear world war actually wouldn't be as bad or even civilization-ending as fanciful sci-fi authors make it out to be. But dismissing its risks entirely because there are some nuclear-war forums where people come up with weird radiation thought experiments just seems preemptive to me.

Re: Licensing is neither feasible nor effective for addressing AI risks

#155

Post by gmerc »

dontupvoteme wrote: post #153
gmerc wrote: post #150

Earlier quoted context omitted.

That’s not how growth works. The moment the current subsidized growth stops the ecosystem would come to a screeching halt.

Growth is not always so simple. The countless GPUs previously used for PoW crypto - Where have they gone?

the GPUs that matter here are A/H100s not randomly 30x consumer cards or mining specific chips

there is a 30x performance difference between generations on AI, the cards from the crypto hypecycle don’t do much anymore for the type of AI work that matters.

Re: Licensing is neither feasible nor effective for addressing AI risks

#156

Post by z3c0 »

pixl97 wrote: post #126
z3c0 wrote: post #97

Earlier quoted context omitted.

While I agree with the other comment, I'd like to add one thing to help you see the false equivalency being made here: we didn't make the human brain . Now, with that being understood, why wouldn't we understand a brain that we made? Don't say "emergent properties", because we understand the ermegent properties of ant colonies without having made them.

You seem to misunderstand the size of the complexity space we are dealing with here. We take simple algorithms feed massive amounts of data to the algorithms, and the algorithm makes a network connecting it together. Humans did not choose anything about how that network is connected... the data and feedback loop did. >because we understand the ermegent properties of ant colonies without having made them. I would disa…

> You seem to misunderstand the size of the complexity space we are dealing with here.

I don't. The whole point of my job is to understand this exact degree of complexity. The rest of your comment rides on that assumption, and is moot. As I said, emergent properties are often unexpected, but once the observation has been made, it is possible to work back through the system to understand how the property came to be (because the system is still deterministic, though not apparently obvious.) Having a billion input features does not change this.

Frankly, understanding the oddness of a ML model is not half as difficult as people make it, and might even be easier than predicting ant colonies. Our inability to predict their behavior has nothing to do with our ability to understand what features influence which behaviors, only that measuring the current state of an ant colony is difficult because it's not like they're all housed within an array of memory. LLMs are, thus we'd understand them better if the most common implementations were not hidden away. Case in point, all the optimization techniques that have emerged from having LLaMa available.

Re: Licensing is neither feasible nor effective for addressing AI risks

#157

Post by z3c0 »

pixl97 wrote: post #132
z3c0 wrote: post #43

Earlier quoted context omitted.

Astroturfing, psyops, and disruption of general trust (commercially and maliciously, both domestic and foreign) It is disturbing to me how unconcerned everybody is with this over what is still only a hypothetical problem. States and businesses have been long employing subversive techniques to corral people towards their goals, and they all just got an alarmingly useful tool for automated propaganda. This is a problem…

I mean it's not hypothetical, there are shit loads of bots on social media already causing problems. Now the bots these days have a human controller but we can see strife they cause in politics already.

We're in accordance.

Re: Licensing is neither feasible nor effective for addressing AI risks

#158

Post by dontupvoteme »

gmerc wrote: post #155
dontupvoteme wrote: post #153

Earlier quoted context omitted.

Growth is not always so simple. The countless GPUs previously used for PoW crypto - Where have they gone?

the GPUs that matter here are A/H100s not randomly 30x consumer cards or mining specific chips there is a 30x performance difference between generations on AI, the cards from the crypto hypecycle don’t do much anymore for the type of AI work that matters.

At the moment, because the current meta is 100B+ Parameter LLMs to do everything. I would not discount chaining specialized smaller models together as a viable approach, and those can run on 24GB or even less.

Re: Licensing is neither feasible nor effective for addressing AI risks

#159

Post by dredmorbius »

pixl97 wrote: post #134
killjoywashere wrote: post #80

Earlier quoted context omitted.

That's actually why it's urgent to set this up outside of government, like the browser CA system, and develop methods to issue verification of "human" while preserving other aspects of anonymity.

There is no outside of government, they hold a monopoly in violence.

The state's monopoly, qua Max Weber, is on the claim to the legitimate use of violence. That is, the right and legitimacy of that right, is restricted to the state, or an entity acting in the effective capacity of a state, whatever it happens to call itself.

Absent this, one of three conditions exist:

1. There is no monopoly. In which case violence is widespread, and there is no state.

2. There is no legitimacy. In which case violence is capricious.

3. Some non-state power or agent assumes the monopoly on legitimate violence. In which case it becomes, by definition The State.

The state's claim is to legitimacy. A capricious exercise would be an abrogation of legitimacy

Weber, Max (1978). Roth, Guenther; Wittich, Claus (eds.). Economy and Society. Berkeley: U. California Press. p. 54.

https://archive.org/details/economysociety00webe/page/54/mod...>

There's an excellent explanation of the common misunderstanding in this episode of the Talking Politics podcast: https://play.acast.com/s/history-of-ideas/weberonleadership>

The misleading and abbreviated form that's frequently found online seems to have originated with Rothbard in the 1960s, and was further popularised by Nozick in the 1970s. It's now falsely accepted as a truth when in fact it is a gross misrepresentation and obscures the core principles Weber advanced.

Re: Licensing is neither feasible nor effective for addressing AI risks

#160

Post by killjoywashere »

dredmorbius wrote: post #159
pixl97 wrote: post #134

Earlier quoted context omitted.

There is no outside of government, they hold a monopoly in violence.

The state's monopoly, qua Max Weber, is on the claim to the legitimate use of violence. That is, the right and legitimacy of that right, is restricted to the state, or an entity acting in the effective capacity of a state, whatever it happens to call itself. Absent this, one of three conditions exist: 1. There is no monopoly. In which case violence is widespread, and there is no state. 2. There is no legitimacy. In w…

Great fact-checking, thank you.
Post reply on HN