Rank the following based on the likelihood to nullify AI risk (whether by achieving alignment, stopping AI development, or another way) If you think you have better solutions to AI risk than I came up with, please add them to your ranking. [1] * Give EY[2] $10M * Give EY...
I recently read What do ML researchers think about AI in 2022. The probability of Doom is sub-10%. Which is high, but as I understand it, in the minds of people like Eliezer Yudkowsky, we're more likely doomed than not. I personally lean towards Yudkowsky's views, because - I don't...
For those wondering about Raistlin Majere, this is from Wikipedia:
« Born to a mother prone to trance-like fits and a woodcutter father, Raistlin inherited his mother's aptitude for magic. He undertook and passed the arduous Test of High Sorcery, but in the process, he acquired white hair and golden skin and was cursed with hourglass eyes which saw the effects of time on all things. His health, while never robust, was ruined further, leaving him weak and subject to frequent bouts of coughing blood. Initially wearing the white robes of good, as the first series progresses Raistlin's powers increase while his mood and actions darken, he goes to neutral red robes for the... (read more)
This was SO INTERESTING. In addition to being insanely brave. Thank you.
was exists
typo / grammatical error
The main advantage for epistemics of working as a moderate is that almost all of your work has an informed, intelligent, thoughtful audience. I spend lots of time talking to and aiming to persuade AI company staff who are generally very intelligent, knowledgeable about AI, and intimately familiar with the goings-on at AI companies. In contrast, as a radical, almost all of your audience—policymakers, elites, the general public—is poorly informed, only able or willing to engage shallowly, and needs to have their attention grabbed intentionally. The former situation is obviously way more conducive to maintaining good epistemics.
This sounds like the streetlight bias, superficially? Just because your audience is intelligent and knowledgeable, doesn't mean it's the right audience, and doesn't mean the stance that led you to them is correct.
Just want to say I absolutely loved the story you wrote with Lintamande between Carissa and Altarrin. Thank you for that 🙏🏻
Idk if I'm the only one here, but I use LLMs for coding and I have disabled memory. This wasn't really an educated move, but after having an AI completely hallucinate in one chat and get the problem or the code at hand totally wrong, I'm afraid its misunderstanding will contaminate its memory and mess up every new chat I start.
At least with no memory I can attempt to rewrite my prompt in a new window and hope for a better outcome. It forces me to repeat myself a lot, but I now have systems in place to briefly summarize what my app does, show the file structure, and explain the problem at hand.
It doesn’t seem to be an uncommon experience that a model is given e.g. a piece of code with a bug in it and asked to find the bug, and then it keeps repeatedly claiming that it “found” the bug and offering revised code which doesn’t actually fix the problem. Or have anything to do with the problem, for that matter.
My personal insanity generator. Sigh.
Thanks for putting this on my radar :)
A few thoughts.
TLDR several reasons to be suspicious of PRT and its creator, but anecdotally it seems to ~work on me.
The negative first:
Super glad I landed on your post! Just ordered the game.
Rank the following based on the likelihood to nullify AI risk (whether by achieving alignment, stopping AI development, or another way)
If you think you have better solutions to AI risk than I came up with, please add them to your ranking.
I recently read What do ML researchers think about AI in 2022.
The probability of Doom is sub-10%. Which is high, but as I understand it, in the minds of people like Eliezer Yudkowsky, we're more likely doomed than not.
I personally lean towards Yudkowsky's views, because
- I don't believe human/evolution-selected minds have thinking power that a machine could not have
- I believe in the Orthogonality Thesis
(I think that those two questions can be defended empirically)
- I think it is easier to make a non-aligned machine than an aligned one
(I believe that research currently being carried out strongly hints at the fact that this is true)
- I believe that more people are working on... (read 893 more words →)
I liked this!
Isn’t this post an elaborate way of saying that today’s posteriors are tomorrow’s priors?
As in- all posteriors eventually get baked into the prior.