I am not really sure what all of the things you are saying here are supposed to tell me. Maybe I am supposed to respect random people in the UN? I do not generally think highly of the UN, or think involvement in it is much of a sign of being a good strategist (though of course, as all highly selected positions it is of course evidence of being in the top percentiles of competence, but not more than that).
I didn't quote these sections because they too are largely uninformative:
Gill helped secure high-impact international consensus recommendations on regulating Artificial Intelligence (Al) in lethal autonomous weapon systems in 2017 and 2018, the draft Al ethics recommendation of UNESCO in 2020, and a new international platform on digital health and Al.
Like, what is this supposed to tell me? I really don't know the sign of lethal autonomous weapon regulation. My guess is it's mildly bad and I was historically opposed to regulating it, but it's not super clear and I've flipped back and forth a few times. The "platform for digital health and AI" seems like a red flag, but I don't know.
1
1
1
1Oh, oops, yep, I confused the two. I meant geek/mop/sociopath in the David Chapman sense. Thank you for the catch!
If one of your central takeaways from AI is that it is "going to help accelerate the process of achieveing the UN's Sustainable Development Goals" then you are deeply miscalibrated about the impact of AI.
It's like saying that "the industrial revolution could help improve the efficiency of chariot production". Bro, there is going to be no chariots after the industrial revolution. There are also going to be no more sustainable development goals post-ASI.
Like, it's a random quote, maybe he had more context that makes it make more sense, but it's the only object level take of his I could find on his Wikipedia page. If he has more relevant things to say, they didn't make it into the things I could quickly find out about him, but my first skim of things strongly suggested someone who lacks situational awareness (in the https://situational-awareness.ai/ sense).
By the way I think I replied to your pro-DIY article. I think a theme between the two is you didn't identify that these things are tradeoffs. At best, you're insightful that people go too much in the other direction, but you didn't articulate what that other direction is, or why they might sometimes do it. Ideally you have something more like a rubric or pros/cons lists for when to go one way or the other.
I think this would be actively bad for a set of internal company principles! Facebook's central motto was "Move fast and break things" not "Move fast and break things in this situation, but not in this other situation". The latter doesn't really work as a principle!
The conditional for all of these principles is "what I think is the right choice for someone working at Lightcone". Much of the force of these principles comes from conditioning on our specific context. The force of a company culture principle comes from all the behaviors that are appropriate in other contexts that it rules out as not being appropriate in this context.
The first thing that immediately comes to mind when I think about how to act in such an environment is reputation: trying to determine which actors are adversarial based on my (or other's) previous interactions with them. I think I would try this before resorting to the other three tactics.
Yeah, totally agree. In general group dynamics around this kind of adversarial stuff are things that I was only able to get a bit into.
That said, I was hoping to include this kind of strategy under the broad umbrella of "You try to purge the untrustworthy". Like, the less intense version of that is to just try to surround yourself with more trustworthy people, and generally increase the degree to which you have good measurements of trustworthiness.
Yep, we've been working on many changes the last few months that drastically sped up how fast the site loads! We have a few more things coming, after which things core content on a page reliably should load in <200ms. Glad to see someone noticed!
8Yeah, this is a big thing that I hope to write more about. Like, a huge dynamic in these kinds of conflicts is someone reinforcing the paranoia, which then elevates more flailing, causing someone's environment to become more disorienting, which causes them to become worse at updating on evidence and become more disoriented, which then makes it easier to throw them off balance more.
Like, as is kind of part of this whole dynamic, the process you use to restabilize yourself in adversarial environments can itself be turned against you.
Amandeep Singh Gill
From Wikipedia:
Gill believes AI can help accelerate the process of achieving the UN's Sustainable Development Goals.[17]
What? I have never heard of this person, and the little I have read suggests he is deeply deeply confused about the nature of AGI. This doesn't feel like a serious suggestion.
1My understanding is that the way it works is that you accept a non-trivial amount of liability even if you say that it's not medical and legal advice, but it usually isn't worth it. Beyond that, many of the world's institutions have limited themselves to only accept "official" legal/medical/financial advice, and so this kind of information has a lot of trouble propagating.
I agree it's not the case you can't talk about these things at all! I might clean up the wording to make that more clear.
I do also think that people suck at being paranoid in the right way, but it's a tricky problem.
I am hoping to write more about how to be paranoid in the right way, or avoid paranoia-inducing environments (my post yesterday was pointing at one such thing), but it's something I am somewhat less confident in than the basic dynamics.