LessWrong 2.0 Reader

I'd love to donate to Lightcone ~5K€ next year, but as long as it's not tax-deductible in France I'll keep to French AI safety orgs as the French non-profit donation tax break is stupidly good: it can basically triple the donation amount and reduce income tax to 0.

I know that Mieux Donner, a new French effective giving org, is acting as French tax-deductible front for a number of EA organisations. I'll contact them to check whether they could forward the donation 

I understand the mechanisms of your metaphor and while they are undeniable they are also, IMO overly simplistic and not helpful as it leaves out the most important element of willpower.

Just like freewill, willpower implies making a decision then using your "freewill" or willpower to act on it. Willpower, as with freewill, is meaningless without making a conscious and preferably well informed decision. 

Regardless of living or dead if you do not have reasonably accurate and somewhat complete info aka less wrong-ish its a crapshoot as to the outcome. 

I posit that humans in general and Americans and most Europeans too have been so thoroughly and consistently lied to that any discussion of willpower or freewill is moot. 

What IS being shut down or weeded out is truth, right and wrong, reality, critical thinking, honesty, and on and on, without which we all loose our ability to make good decisions.

I don't know how our species can overcome this but IOM nothing else matters until we do.

Agreed, I think people should apply a pretty strong penalty when evaluating a potential donation that has or worsens these dynamics. There are some donation opportunities that still have the "major donors won't [fully] fund it" and "I'm advantaged to evaluate it as an AIS professional" without the "I'm personal friends with the recipient" weirdness, though -- e.g. alignment approaches or policy research/advocacy directions you find promising that Open Phil isn't currently funding that would be executed thousands of miles away.

if brute force doesn’t solve your problem, then you are not using enough.

Stealing that!

Although not much, I have donated 10$. I hope you will find some generours sponsors that are able financially support Lightcone!

Actually never mind. But for future reference I guess I’ll use the intercom if I want an old version labeled. Thanks for telling me how that works.  :)

(There’s a website / paper going around that cites a post [LW · GW] I wrote way back in 2021, when I was young and stupid, so it had a bunch of mistakes. But after re-reading that post again this morning, I decided that the changes I needed to make weren’t that big, and I just went ahead and edited the post like normal, and added a changelog to the bottom. I’ve done this before [LW(p) · GW(p)]. I’ll see if anyone complains. I don’t expect them to. E.g. that same website / paper cites a bunch of arxiv papers while omitting their version numbers, so they’re probably not too worried about that kind of stuff.)

My answer is basically yes.

I was only addressing the question "If we basically failed at alignment, or didn't align the AI at all, but had a very small amount of niceness, would that lead to good outcomes?"

I'm someone who doesn't have children and doesn't plan to, but I agree with pretty much everything in this post. I do have reservations on how to go about justly raising someone's taxes now because of now-irreversible decisions they made not to have children in the past before the policy existed, but am all in favor of lump sums paid by future taxes, including higher taxes on me personally.

Is it sad I have higher expectations that society will solve the really hard technological problem of aging in the next couple of decades than the economically-should-be-straightforward problem of fertility gaps?

My take - lots of good analysis, but makes a few crucial mistakes that throw the conclusions into significant doubt:

The USG will be able and willing to either provide or mandate strong infosecurity for multiple projects.

I simply don't buy that the infosec for multiple such projects will be anywhere near the infosec of a single project because the overall security ends up being that of the weakest link.

Additionally, the more projects there are with a particular capability, the more folk there are who can leak information either by talking or by being spies.

The probability-weighted impacts of AI takeover or the proliferation of world-ending technologies might be high enough to dominate the probability-weighted impacts of power concentration.

Comment: We currently doubt this, but we haven’t modelled it out, and we have lower p(doom) from misalignment than many (<10%).

Seems entirely plausible to me that either one could dominate. Would love to see more analysis around this.

Reducing access to these services will significantly disempower the rest of the world: we’re not talking about whether people will have access to the best chatbots or not, but whether they’ll have access to extremely powerful future capabilities which enable them to shape and improve their lives on a scale that humans haven’t previously been able to.

If you're worried about this, I don't think you quite realise the stakes. Capabilities mostly proliferate anyway. People can wait a few more years.

Update: I'm now familiar with the term "demand avoidance". One recommendation for caregivers is "declarative language". On LessWrong we might call it "guess culture [LW · GW]" or perhaps "tell culture [LW · GW]". Aesthetically I dislike it, but it works for this child (in combination with other things, including your good advice of persuasion and positive reinforcement).