LessWrong 2.0 Reader

The claim is that most applications aren't internal usage of AI for AI development and thus can be made trivially safe.

Not that most applications of AI for AI development can be made trivially safe.

What do you think of this post by Tammy?

It seems like someone could definitely be wrong about what they want (unless normative anti-realism [LW · GW] is true and such a sentence has no meaning). For example consider someone who thinks it's really important to be faithful to God and goes to church every Sunday to maintain their faith and would use a superintelligent religious AI assistant to help keep the faith if they could. Or maybe they're just overconfident about their philosophical abilities and would fail to take various precautions that I think are important in a high-stakes reflective process.

Mostly that thing where we had a lying vs lie-detecting arms race and the liars mostly won by believing their own lies and that’s how we have things like overconfidence bias and self-serving bias and a whole bunch of other biases.

Are you imagining that the RL environment for AIs will be single-player, with no social interactions? If yes, how will they learn social skills? If no, why wouldn't the same thing happen to them?

Unless we do a very stupid thing like reading the AI’s thoughts and RL-punish wrongthink, this seems very unlikely to happen.

We already RL-punish AIs for saying things that we don't like (via RLHF), and in the future will probably punish them for thinking things we don't like (via things like interpretability). Not sure how to avoid this (given current political realities) so safety plans have to somehow take this into account.

While I've softened my position on this in the last year, I want to give a big +1 to this response, especially these two points:

• It's genuinely hard to come up with ideas that help capabilities a lot. I think you are severely underestimating how hard it is, and how much insight is required. I think one issue here is that most papers on arxiv are garbage and don't actually make any progress, but those papers are not the ones that are pushing AGI forward anyways.
• [..]
• High level ideas are generally not that valuable in and of themselves. People generally learn to ignore ideas unless they have strong empirical evidence of correctness (or endorsement of highly respected researchers) because there are simply too many ideas. The valuable thing is not the idea itself, but the knowledge of which ideas are actually correct.

(emphasis added)

I think it's often challenging to just understand where the frontier is, because it's so far and so many things are secret. And if you're not at a scaling lab and then also don't keep up with the frontier of the literature, it's natural to overestimate the novelty of your insights. And then, if you're too scared to investigate your insights, you might continue to think that your ideas are better than they are. Meanwhile, as an AI Safety researcher, not only is there a lot less distance to the frontier of whatever subfield you're in, you'll probably spend most of your time doing work that keeps you on the frontier. 

Random insights can be valuable, but the history of deep learning is full of random insights that were right but for arguably the wrong reasons (batch/layernorm, Adam, arguably the algorithm that would later be rebranded as PPO), as well as brilliant insights that turned out to be basically useless (e.g. consider a lot of the Bayesian neural network stuff, but there's really too many examples to list) if not harmful in the long run (e.g. lots of "clever" or not-so-clever ways of adding inductive bias). Part of the reason is that people don't get taught the history of the field, and see all the oh-so-clever ideas that didn't work, or how a lot of the "insights" were invented post-hoc. So if you're new to deep learning you might get the impression that insights were more causally responsible for the capabilities advancements, than they actually are. Insofar as good alignment requires deconfusion and rationality to generate good insights, and capabilities does not, then you should expect that the insights you get from improving rationality/doing deconfusion are more impactful for alignment than capabilities. 

I mean, if you actually do come up with a better initialization scheme, a trick that improves GPU utilization, or some other sort of cheap algorithmic trick to improve training AND check it's correct through some small/medium-scale empirical experiments, then sure, please reconsider publishing that. But it's hard to incidentally do that -- even if you do come up with some insight while doing say, mech interp, it feels like going out of your way to test your capability ideas should be a really obvious "you're basically doing capabilities" sign? But maybe, you should be doing your safety work instead?

Retinoids, which is a big family of compounds but I would go with adapalene, which has better safety/side effect than anything else. It has less scientific evidence for anti-aging than other retinoids (and is not marketed for that purpose), but I've tried it myself (bought it for acne), and it has very obvious anti-wrinkle effects within like a week. You can get generic 0.1% adapalene gel on Amazon for 1.6oz/$12.

(I'm a little worried about long term effects, i.e. could the increased skin turnover mean faster aging in the long run, but can't seem to find any data or discussion about it.)

Regarding Cortez and the Aztecs, it is of interest to note that Cortez's indigenous allies (enemies of the Aztecs) actually ended up in a fairly good position afterwards.

From https://en.wikipedia.org/wiki/Tlaxcala

For the most part, the Spanish kept their promise to the Tlaxcalans. Unlike Tenochtitlan and other cities, Tlaxcala was not destroyed after the Conquest. They also allowed many Tlaxcalans to retain their indigenous names. The Tlaxcalans were mostly able to keep their traditional form of government.

It implies that AI control is organizationally simpler, because most applications can be made trivially controlled.

I didn't get this from the premises fwiw. Are you saying it's trivial because "just don't use your AI to help you design AI" (seems organizationally hard to me), or did you have particular tricks in mind?

Finding a loan to move to somewhere with jobs is probably your best bet. This may devolve to begging amongst any social circles as well, which is a big pride hit. Many probably won't believe you will wind up with the means to pay it back. Minimizing the cost of the move by getting rid of non essential belongings. It is probably somewhat easier these days to line up a far away job via zoom interviewing. Quantity over quality.

One thing that I think is non-obvious: if you lay out the case for the loan in detail, that demonstrates intelligence and conscientiousness and will increase people's sense that you are doing something useful and thus willingness to lend. Basically, treat it as a business case, but the business happens to just be getting you and your partner employed. Show past cash flow, expected future cash flow given salaries of the sorts of places you are applying to, estimates of how many jobs you can reasonably apply to, how much of a monthly payment you could pay back, etc. Use chatgpt for help with outlining this.

'Evidence based' in the skincare industry mostly means a company paid a consultancy to do 'scientific testing.' Very little shows any signs of actual large effect sizes other than blocking UV damage, afaik. Moisturizer does seem to help, especially in low humidity climes. Which moisturizer your skin responds best to seems to be trial and error, ie the detectable/subjective metrics are all we really have. Retinoids are one exception, but come with tradeoffs that I don't fully understand.

I settled on snail mucin as a more natural alternative, but a significant fraction of people have mild allergic reactions to it apparently.

[epistemic status: I think I’m mostly right about the main thrust here, but probably some of the specific arguments below are wrong. In the following, I'm much more stating conclusions than providing full arguments. This claim isn’t particularly original to me.]

I’m interested in the following subset of risk from AI:

• Early: That comes from AIs that are just powerful enough to be extremely useful and dangerous-by-default (i.e. these AIs aren’t wildly superhuman).
• Scheming: Risk associated with loss of control to AIs that arises from AIs scheming
  • So e.g. I exclude state actors stealing weights in ways that aren’t enabled by the AIs scheming, and I also exclude non-scheming failure modes. IMO, state actors stealing weights is a serious threat, but non-scheming failure modes aren’t (at this level of capability and dignity).
• Medium dignity [EA · GW]: that is, developers of these AIs are putting a reasonable amount of effort into preventing catastrophic outcomes from their AIs (perhaps they’re spending the equivalent of 10% of their budget on cost-effective measures to prevent catastrophes).
• Nearcasted: no substantial fundamental progress on AI safety techniques, no substantial changes in how AI works.

This subset of risk is interesting because I think it’s a natural scenario at which to target technical work on AI safety. (E.g. it’s the main scenario we’re targeting with our AI control agenda [LW · GW].)

I claim that the majority of this risk comes from the lab using AIs internally to do AI development (by which I mean both research and engineering). This is because the AIs doing AI development naturally require access to compute and model weights that they can potentially leverage into causing catastrophic outcomes—in particular, those resources can be abused to run AIs unmonitored.

Using AIs for AI development looks uniquely risky to me among applications of early-transformative AIs, because unlike all other applications I know about:

• It’s very expensive to refrain from using AIs for this application.
• There’s no simple way to remove affordances from the AI such that it’s very hard for the AI to take a small sequence of actions which plausibly lead quickly to loss of control. In contrast, most other applications of AI probably can be controlled just by restricting their affordances.

If I’m right that the risk from scheming early-transformative models is concentrated onto this pretty specific scenario, it implies a bunch of things:

• It implies that work on mitigating these risks should focus on this very specific setting.
• It implies that AI control is organizationally simpler, because most applications can be made trivially controlled.
• It is pretty plausible to me that AI control is quite easy, because you actually can remove affordances from the AIs that are doing AI R&D such that it’s hard for them to cause problems.

This is a great question, and one of the things I'm most excited about using this framework to study in the future! I have a few ideas but nothing to report yet.

But I will say that I think we should be able to formalize exactly what it would mean for a transformer to create/discover new knowledge, and also to apply the structure from one dataset and apply it to another, or to mix two abstract structures together, etc. I want to have an entire theory of cognitive abilities and the geometric internal structures that support them.