LessWrong Team
I have signed no contracts or agreements whose existence I cannot mention.
Ruby
Karma: 16,415
So I publicly wrote that I was throwing my agency at my problems this year, here’s an interim update:
Attempted to send some longitudinal biosamples[1] to a research label, but shipping was delayed and dry-ice ran out, so they thawed and I lost the batch. Recollecting and trying again.
Nearly weaned off Lexapro. It relieved a certain kind of anxious feeling but didn’t address primary anticipatory insomnia, and had side effects I didn’t like.
Tried ketamine (sublingual, dose matched to clinical anti-depression 40-minute IV dose, then divide by four to account for my impaired drug metabolism). Resulted in heavy sedation/dizziness + elevated mood, long-lasting – unpleasant but it did abort the bipolar down-state (possibly even swung a bit into elevated mood. I’ll probably try again when I next have a bipolar downstate, but with half the dose again.
Tried Gabapentin once for anticipatory insomnia (pre-motorsport event). Low-dose, seemingly effective and sufficiently out of my system by morning. (Gabapentin is excreted renally and bypasses my weird hepatic drug metabolism.
I’m now regularly taking fish oil, vitamin D3, and magnesium. Just started on NAC today.
Gabapentin and ketamine seem promising. Lexapro was information but not a long-term solution in and of itself. The recent period has been pretty stable, but hard to attribute to e.g. supplements at this time.I might be forgetting some other stuff, and also I don’t think I gave the list of everything else from earlier in this year, e.g. parasympathetic activation interventions (vagus nerve stimulation, HRV biofeedback), bright light in mornings, blue-blocking glasses for the evening, and other stuff.
- ^
I’m not going to give more detail.
Can you say more about what process you’re imagining here?
If I see another person complaining that no one reads their post when it starts with a 15-sentence epistemic status meta-commentary that doesn’t give me any reason to want to keep reading, I will explode into a violent burst of fury and laser beams.
Oh yeah, this one is definitely about me.
In the opposite direction, I think the Nuclear winter fears were inflated by Sagan and others, for political purpose, but those fears have endured to this day.
Thank you!
I think if your time/money budget allows, then nothing says you can’t get them something else...
Perishable items feel like they fall on some spectrum between non-perishable items and experiences, and I’m already in favor of experiences as being good gifts. I think your subconscious is just flat wrong: if you go out and look for premium chocolates, flowers, art supplies, or anything, really, there are very expensive premium ones that very often are actually better.
I listened to a podcast once about a guy who had a predatory loan shark empire and repented and gave it up for moral reasons.
Look, Sarah Chen is anthropically a very plausible name to encounter.
It’s an always-on foreground service. Indeed, works from lock-screen.
@Neel Nanda Not software you can try, but a detailed description: My Exobrain Software (forays into cyborgism)
Thanks! Sounds good. Yeah, I’ll check for those variants.
Regarding PRS quality, indeed. There’s a table in a collapsible section with an analysis of the quality of PRSs used. Interesting regarding your own conversion from GWAS to PRS.Your VCF is not enough to calcualte your PRS, because the reference (rather than variant) is the “effect allele” for many scores of interest; the default behaviour of tools like PRSKB CLI and PSG_Calc is to impute the missing alleles based on your VCF but this is going to be wrong, you likely want to re-call variants from your CRAM (this is dicussed at length in the blog post I linked earlier)
Ah, cool, yes. Interestingly Claude/GPT left a comment in its code mentioning exactly this problem, and then punted on it and I didn’t notice.
CNVs are fairly tractable to calculate from CRAMs
We attempted this and it failed because of contig mismatch with the reference on the CRAM. Going back to it, we could have just downloaded the appropriate one? (DRAGEN/Lumina?) Another thing not done for no good reason that I didn’t catch. (Other things I did catch, but not this.)
Cool, that gives me some things to do.
(To clarify, I don’t think that I have default poor executive function, but that it is still possible to have yet more. I do think the ups and downs of bipolar make habits harder to establish.)
I don’t have all my logs fit into the context window, they’d never fit. I ask it to load things when relevant, and handle things via search or other scripts for operating over the past. Todo items is more relevant. They technically fit but it’s costly to have a full-context window so I preload only recent and high-priority ones.
My system does have a bunch of offline processing where raw transcripts get mined and parsed in other documents. In one case, there’s two layers of that.
I’ve been doing a lot of large research projects with LLMs recently. A project might involve 10-20 “Claude Research Mode” runs or equivalent analyses with code, where each one is maxing out context.
The key thing is “session handoff” and “subagent calls”. I do all of this by hand, and I avoid having the harness automatically compact conversations (I don’t like opaque compression, I want to know what’s in context for any given chat instance.)
Whether or not I’m doing it well, I think there’s metis in where and how I split up the work (and have the model split up the work for itself), and for research this is worth working on.
I don’t have the perfect process up yet. But it’s definitely the case there are research projects worth tackling that are 20x as big as context windows, and LLMs are useful, and it’s not a bad process to be involved in the process connect the convos and calls together.
I’d expect the LLM to raise the considerations that matter most, but those are:
Too much oxygen is also a problem. You can both under-dose and overdose.
If you lack experience, you might miss cases where you ought to be going to the ER, and instead you’re self-treating poorly at home.
If you could find a doctor who informally signed off on self-administering, or you really really trust you research skills, then maybe try it, but I’d caution against.
Thanks for the great comment. In my mind, the situation is something like Anthropic is pausing but also hoping to resume eventually, and continues its general approach of saying less rather than more about strategy and what’s going on internally. If that were the case, though, maybe they’d just not do further training runs internally without advertising this fact. One guess is preventing leaks is hard and safety-minded (quite senior) employees might be forcing action, though I admit the pause requires better motivation than the story provides.
Regarding communicable evidence: on the one hand, there are public statements about pausing going along with compelling shareable evidence, but on the other, I think historically it’s just been a pretty secretive company. Many of my conversations with Anthropic employees hit “I can’t talk about that.” It feels hard to see that changing.Even in the story, it’s not clear the world is on track for meaningful global governance of the kind that would stop OAI/GDM though.
Completely. The point of the story was not to imply that the pausing results in everything working out great, merely that pausing would be a Highly Significant Event with Some Pretty Large Effects. Adequate legislation conditional on significant global legislation efforts feels like < 50% to me, at least.
Parts of the story that are now feeling implausible to me in the reaction are (a) the rapidity of reaction, especially the joint statement, which, if it did happen, would probably take much longer, (b) the conjunction of all those things happening – I think any of them individually or a subset is plausible.My sense is that China mostly doesn’t believe in x-risk and mainly wants to be freed from US export controls so it can compete fairly,
That seems plausible. Mostly, it seems no one is trying to talk to China and take it for granted that they’ll want to compete and not take x-risk seriously enough to sign a treaty. I’m curious about your sources/evidence here.
Meanwhile, safety research at Anthropic basically collapses because they can no longer afford compute
I’m curious what the cost of ongoing safety research with existing models is if you’re just doing inference. My guess is not nearly as much as the big training runs? Other variables are how long do their products stay frontrunners, and are they profitable vs subsidized by investment still at this point. But that’d be like 6-12 months max I’m guessing. I don’t have the best models here.
It’s something I built. I think theoretically I should be fine open-sourcing it, but it’s not very turnkey and probably 6-12 (??) hours of setup for someone else to use (like setting up DB, Vercel, modifying prompts). Perhaps at some point I’ll get it into a state where it’s not like that.
This is a good comment and the kind of discussion I hoped this post would start.
I find what you say pretty compelling that the nature of the pause isn’t actually realistic to how Anthropic would behave.
To defend the piece though the goal wasn’t to answer “conditioning on an Anthropic pause, what would that pause most likely look like?” and instead “conditioning on a pause, how would the world react?”
The intention is to respond to the claim that unilaterally pausing would be foolish as worse actors would simply move ahead. I think that’s really far from obvious. A unilateral pause would have a huge impact. It could be debated whether that’s better than winning, but I think it’s crazy to speak as though a unilateral pause would do definitely do nothing. If this piece feels at all believable conditioning on a pause (e.g. what would happen if Dario acted like a LW-reader), then it’s succeeded at my goal.
Aside; the lack of mention of the Anthropic/DoW conflict is a hole in the post.
Done!
Please sir, mercy. I’m in Pacific Time and it was 11pm here!
Curated. A lot could be said about this post and a lot is being said about this post (see the 111 comments on it), but a thing I find neat about it is that it continues the debate around an old idea that’s still important and perhaps cruxy for many.
Oneshotness is not new (not a new concept, not a new argument, not a new debate); yet it is a critical argument for alignment difficult, and is still contested, with implications for what people do. Having made many attempts to explain their side, I could imagine someone giving up and being like “the people who get it, get it, and those who don’t, don’t” and not bothering further to try to argue it. I lean that way. Eliezer admits frustration and the tone matches, but he’s still talking and writing with the hope of being convincing. People are responding, debating further. It’s heated and we’re in the realm of accusations of motivated cognition, strawmanning, motte & bailey, but in this post Eliezer is bringing object level arguments (e.g. analogies) in attempt to advance the conversation, and 111 comments later, conversation is being had. It might not be enough, but I think there are some dignity points in that.