Masters Degree in Biotechnology. Work in Pediatric Hematology Research.
HiddenPrior
HiddenPrior’s Shortform
Both here, and when the story was originally posted, it seems that most of the disagreements or objections are focused on the epistemic question of whether or not The Owned Ones are moral patients, or if the Owners arguments are straw men and might actually be right, etc.
I am pretty confident this is missing Eliezer’s point: The point of the story is that these objections are fairly standard and they are the weak arguments of someone who does not care to investigate the issue far enough to make a more sophisticated argument.
The problem isn’t that the owners are necessarily wrong, it is that they clearly don’t care.
Hopefully that helps dispel some of the confusion regarding the stories purpose.
Hmmm, as you say it is a fairly unambitious story. IMV the purpose of the story is simply to give an alternative perspective to the most standard and basic dismissals of concern for model well being.
As Eliezer points out at the end, the issue isn’t that the owners are necessarily wrong; this isn’t meant to address that. It is simply pointing out that the way the owners behave is unacceptable because it is how you act if you simply don’t care.
Trying to stop the development of AI today, is kind of like trying to have stopped the manhattan project.
You are correct that stopping the Manhatten project would have been much easier.
It was a secret project, and the incentives for most of the decision-makers involved were around producing “better-but-accountable” outcomes. There were multiple individuals to whom if you presented a sufficiently plausible argument, might have been able to stop the Manhattan project, up until the actual attack on Hiroshima. Fermi et al. had a pretty good chance of preventing the bomb from being used, and had other technical experts agreed that the risks were too high, I think the Manhattan project could have been stopped.
The current situation is much worse, with far more people having strong incentives to deceive others, themselves, and knowingly take risks.
All of that said, I still don’t think this means we have no choice. There are multiple ways of coordinating to solve a bad equilibrium, and I don’t think the incentives for pursuing AI are nearly so universal.
Most of the public recognizes that they have a lot of incentive to stop the development of ASI, even if right now it is more focused on job losses, they also recognize that the gains of AI are unlikely to be distributed evenly, and most Americans in particular care more about their chance of gaining in social status, than almost any other terminal outcome of policy (I can provide some citations if needed.)
More importantly, that public disapproval likely has a stable equilibrium in that they probably should be worried about the existential risk, whatever percentage it really is.
All that needs to follow is that the public disapproval needs to be strong enough and translate into a strong enough signal that policy makers are duly incentivize to take the concerns seriously, and policy make.
That isn’t super easy, but it isn’t pure fantasy at all. Everyone can see how the incentives for AI firms favor total race dynamics, but for most of humanity and even most Americans incentives are more than strong enough to prefer caution.
This is very heartening for those of us who are too young or too low on resources to make contributions on short timelines in any situation except one of the most acute desperation; if a college student were to do all they could to stop ASI they could throw their life/career away to try and make a difference, but they only get one shot at that, so it better count or happen when things actually matter that much.
Otherwise, it is good to note that we can still contribute to hedging against long term scenarios by investing in those outcomes. I am going to think deeply on what long/medium term strategies could be the most impactful from my position.
Ya, I agree this should be true in principal; I think given more time, there might be the opportunity for some sort of “Dath Ilan” lite society to rise to the top.
I think this is true in so far as there is selection pressure, in that such events are survivable, and in that they don’t require unified coordination of all agents to survive.
The cthulu example isn’t great, only in that the nature of the threat is pretty vague to most readers (at least to me).
A better example would be; a medieval society gets hit by a meteorite; does this cause selection pressure for medieval societies to build meteorite-proof castles? Not if it just kills everyone.
Alternatively; an early-industrial era society that notices an approaching comet might be able to coordinate to invent a redirect-rocket, or nukes, or whatever to save their society, but there is still no selection pressure since the two possible results are everyone survives, or everyone does, and if anything competitive pressure will punish anyone who spends resources on saving the world, since those resources will benefit competitors who spend zero resources on the asteroid redirect mission, just as much as the ones who spent half their GDP to survive. Unless social pressure or similar can effectively reward the heroic resource sacrificing nations, they will be putting themselves at a huge disadvantage and if GDP correlated to representation over time you would expect the selfish nations to actually be the ones that are selected for.
I get that you arguing that a society which was this bad at reasoning ing general should be outcompeted by a society that is better at reasoning, but we should expect both societies to be out competed by one that is both capable of reasoning well when it is competitively valuable, and ignoring such reasoning when it is not competitively valuable. I think might be a good description of the current United States, for example which is great at listening to academics when it is profitable and ignoring them when it is inconvenient for business interests.
I kept expecting the article to pivot into how this illustrated some fascinating principal of how to reduce catastrophic risk. Was very pleased to be wrong. What a wonderful post.
So exciting to see this essay — I really appreciate the time and quality you put into it.
I currently work as a wet lab manager for an academic biotechnology lab, and I’ve been experimenting with significant workflow changes since setting up an OpenClaw-based agent on an old laptop I had sitting in a closet. The results have been phenomenal. I may write a longer post about this, but briefly: running OpenClaw with Opus 4.6 has been a massive productivity accelerator. I can confidently say I’m getting 2–3x more done per week after just two weeks of adoption.
I was previously unfamiliar with lab automation as a field, and outside of a few niche applications — like automating sample intake for analytical instruments like flow cytometers or mass specs — I would have considered it premature to call lab automation broadly useful anytime soon. That was true until I tried OpenClaw.
Before explaining why I think OpenClaw and recent LLMs change things, I want to highlight what I see as the crux of the challenge: you’re right to point to scale as the limiting factor. The current paradigm in university and most commercial labs is built around minimizing scale, not maximizing it. The objective is to test hypotheses using as few resources as possible. If something works, you scale up to manufacturing, commercial testing, or clinical trials — and those workflows are very automatable, but aren’t really “lab work.”
Any good wet lab scientist lives by something like: success (in the lab) is 80% preparation, 15% documentation, and only 5% execution. Nobody bothers automating most lab work because execution time is rarely the bottleneck or the expensive problem to solve. Most scientists would rather do the bench work themselves just to ensure the protocol is actually followed correctly. The real overhead lives in writing grants, designing meticulous experimental plans, translating those plans into reality, and troubleshooting unexpected blockers.
I expect that reality to change rapidly with AI. Setting up a lab management system with Claude Code and folding it into an OpenClaw setup has dramatically reduced the overhead of preparation. I no longer have to manually type each protocol or the small adjustments needed for slightly different experiments. Instead of spending my last hour each day writing up my lab journal, I dictate specific details to my agent while I’m still in the tissue culture hood — and by the time I’m wrapping up, a well-written, wiki-linked, detailed journal entry is ready for me to review and approve.
The most striking moment was when it started successfully generating cloning protocols and working Python scripts autonomously.
I think cloud-based labs are well-positioned to take off, and you’re right to point toward companies like Plasmidsaurus as the most likely intermediaries between now and fully automatable labs. Plasmidsaurus has completely superseded our university sequencing core in cost, speed, and turnaround time. Services like Genewiz and IDT are increasingly beating our primer synthesis core as well — and now whenever we set up a cloning workflow, the first question is “should we outsource this?”
In our lab, individual workflows can be mapped on a spectrum of automatability. Thermocyclers are cheap and simple enough that everyone has that automation in-house. Everyone outsources gBlocks, sequencing, and increasingly analysis. Cloning sits at the current borderline — many labs now outsource it if the desired construct is standard enough, but specialist needs remain, and the reliability, cost, and turnaround for outsourced cloning is right on the margin of being worthwhile for noncommercial labs.
Tissue culture-dependent workflows are still a ways off. Our lab generates lentivirus, which we test in vitro before moving to in vivo work. Outsourcing lentiviral production exists, but the limited capacity for highly reliable production is entirely consumed by clinical demand and likely will be for the foreseeable future. The combination of rigor and specialist optimization required for lentivirus makes it far more resistant to automation than bacterial workflows, except in fully standardized manufacturing contexts.
I’d still expect lentiviral production to become automatable eventually, and after that, in vivo testing may be partially addressed through organoids or similar approaches. LLMs will dramatically accelerate both the development of these technologies and their diffusion into biotechnology practice broadly.
Super helpful! Thanks!
I am limited in my means, but I would commit to a fund for strategy 2. My thoughts were on strategy 2, and it seems likely to do the most damage to OpenAI’s reputation (and therefore funding) out of the above options. If someone is really protective of something, like their public image/reputation, that probably indicates that it is the most painful place to hit them.
I knew I could find some real info-hazards on lesswrong today. I almost didn’t click the first link.
Same. Should I short record companies for the upcoming inevitable AI musician strike, and then long Spotify for when 85% of their content is Royalty free AI generated content?
I did a non-in-depth reading of the article during my lunch break, and found it to be of lower quality than I would have predicted.
I am open to an alternative interpretation of the article, but most of it seems very critical of the Effective Altruism movement on the basis of “calculating expected values for the impact on peoples lives is a bad method to gauge the effectiveness of aid, or how you are impacting peoples lives.”
The article begins by establishing that many medicines have side effects. Since some of these side effects are undesirable, the author suggests, though they do not state explicitly, that the medicine may also be undesirable if the side effect is bad enough. They go on to suggest that Givewell, and other EA efforts at aid are not very aware of the side effects of their efforts, and that the efforts may therefore do more harm than good. The author does not stoop so low as to actually provide evidence of this, or even make any explicit claims that could be checked or contradicted, but merely suggests that givewell does not do a good job of this.
This is the less charitable part of my interpretation (no pun intended), but I feel the author spends a lot of the article constantly suggesting that trying to be altruistic, especially in an organized or systematic way, is ineffective, maybe harmful and generally not worth the effort. Mostly the author does this by suggesting anecdotal stories of their investigations into charity, and how they feel much wiser now.
The author then moves on to their association of SBF with Effective Altruism, going so far as to say: “Sam Bankman-Fried is the perfect prophet of EA, the epitome of its moral bankruptcy.” In general, the author goes on to give a case for how SBF is the classic utilitarian villain, justifying his immoral acts through oh-so esoteric calculations of improving good around the world on net.
The author goes on to lay out a general criticism of Effective Altruism as relying on arbitrary utilitarian measures of moral value, such as what counts as a life saved. The author suggests Effective Altruism has become popular because Billionaires like how it makes a straightforward case for converting wealth into moral good, and generally attempts to undermine this premise.
The author is generally extremely critical of EA, and any effort at organized charity, and suggests that the best alternative to EA (or utilitarian moral reasoning in general, I presume) is the following:
the “dearest test.” When you have some big call to make, sit down with a person very dear to you—a parent, partner, child, or friend—and look them in the eyes. Say that you’re making a decision that will affect the lives of many people, to the point that some strangers might be hurt. Say that you believe that the lives of these strangers are just as valuable as anyone else’s. Then tell your dearest, “I believe in my decisions, enough that I’d still make them even if one of the people who could be hurt was you.”
Or you can do the “mirror test.” Look into the mirror and describe what you’re doing that will affect the lives of other people. See whether you can tell yourself, with conviction, that you’re willing to be one of the people who is hurt or dies because of what you’re now deciding. Be accountable, at least, to yourself.
Which I suppose is fine, but I think this reveals the author is primarily concerned about their personal role or responsibility in causing positive or negative moral events, and that the author has very little regard for a consequentialist view of the actual state of reality. Unfortunately, the author does very little do directly engage in dialogue about moral values, and makes the assumption throughout the entire article that everyone does, or at least should, share their own moral values.
The author finishes the article with an anecdote of their friend, who they suggest is a better example of being an altruist since they fly out to an island themselves, where they provide direct aid with water stations, and the direct accountability and lack of billionaires demonstrates how selfless and good he is.
I don’t know who this author is, but I get the feeling they are very proud of this article, and they should surely congratulate themselves on spending their time, and the time of their readers so well.
TL;DR
All in all, I think this article can best be summarized by honestly expressing that I feel I wasted my time reading it, and writing this summary. I considered deleting my post on this article, so that I would not risk others also wasting their time on it, but I will leave this summary up so that they can at least waste less time on this article.
Unsure if there is normally a thread for putting only semi-interesting news articles, but here is a recently posted news article by Wired that seems.… rather inflammatory toward Effective Altruism. I have not read the article myself yet, but a quick skim confirms the title is not only to get clickbait anger clicks, the rest of the article also seems extremely critical of EA, transhumanism, and Rationality.
I am going to post it here, though I am not entirely sure if getting this article more clicks is a good thing, so if you have no interest in reading it maybe don’t click it so we don’t further encourage inflammatory clickbait tactics.https://www.wired.com/story/deaths-of-effective-altruism/?utm_source=pocket-newtab-en-us
I am so sad to hear about Vernor Vinge’s death. He was one of the great influences on a younger me, on the path to rationality. I never got to meet him, and I truly regret not having made a greater effort, though I know I would have had little to offer him, and I like to think I have already gotten to know him quite well through his magnificent works.
I would give up a lot, even more than I would for most people, to go back and give him a better chance at making it to a post-singularity society.
“So High, So Low, So Many Things to Know”
I’m sorry you were put in that position, but I really admire your willingness to leave mid-mission. I imagine the social pressure to stay was immense, and people probably talked a lot about the financial resources they were committing, etc.
I was definitely lucky I dodged a mission. A LOT of people insisted if I went on a mission, I would discover the “truth of the church”, but fortunately, I had read enough about sunk cost fallacy and the way identity affects decision-making (thank you, Robert Caldini) to recognize that the true purpose of a mission is to get people to commit resources to the belief system before they can really evaluate if they should do so.
Oh, haha, ya, I didn’t try to convince my parents either, they (particularly my dad) just insisted on arguing as thoroughly as possible about why I didn’t believe in the church/god. Exactly. It says everything about the belief system, when if you ask your parents (which I did) what evidence would convince them to leave, and they say literally no evidence would convince them. I asked, even if God appeared in front of you and said everything except baptism for the dead is true, you wouldn’t believe him? And he insists God would only do that through his prophet, so he would dismiss it as a hallucination lol.
At least for me, dating was a very rocky road after initially leaving the church. Dating in Utah was really rough, and because I was halfway through my undergraduate degree, I wasn’t yet willing to leave. There are a lot of really bad habits of thought and social interaction that the church engrains in you, around social roles and especially shame around sex. Personally, I oscillated heavily between periods of being extremely promiscuous and dating/sleeping with as many people as possible and periods of over-romanticizing and over-committing to a relationship. I think this is normal, but the absence of any sort of sex in my relationships until I was 18 kind of gave me a late start, and my conflicting habits and feelings made things a little crazy.
I did end up getting married very young, in an ill-advised relationship, where the truth is I was trying to please my parents and extended family. I had been dating her for a couple of years and had lived together for more than a year, and the truth is I had a lot of shame about that and wasn’t willing to tell my extended family because my parents were so embarrassed and thought it was such a dark and terrible secret. In the end we divorced after a very short period of time, with my only regret being that we didn’t end things much sooner.
I eventually met someone who was a much better person and who I see as a likely life partner. We have been together for three years now, and our relationship is the best I have ever had and is considerably better than my previous estimates of how fulfilling, enjoyable, and stable a relationship could be. It helps that she is much smarter than me, and we have both learned a lot of lessons the hard way.
My advice as far as dating goes is to not rush into anything. It is so easy because of the social norms in Utah, and the expectations we were raised with within mormonism to feel pressure to get into a relationship, and push that relationship to a very high level of commitment very quickly. In my opinion, the relationship will be healthier, and you are more likely to find a correct one if you tap the breaks as frequently as possible, since you are likely to tend too far in the accelerationist side of the spectrum, especially if you are new to dating. Personally I thought I did a lot of casual dating, but there is a big difference between casual hook ups and actually dating to find a partner, and I think it is important to not conflate what you are really after when you go on dates. I definitely struggled with this.
As far as actually meeting people, this is the main reason it is so important to be slow to form commitments…. I like Scott Alexander’s idea of “micromarriages” as a way to gauge how effective different activities might be at helping you find a good long term relationship. The simple advice though is too avoid dating apps altogether, unless you are just looking to hook up, in which case they are fine, but meeting people in person will still probably lead to a higher quality experience. My own experience, meeting my partner on campus by chance, may skew my perception about what the best way to meet people is, but I really feel that generally people I met in person resulted in better outcomes for my dating life.
The best method is probably to find social events/spaces where people who share your values are likely to attend. Classes can be fine, depending on where you are in Utah, but better are specific social events or clubs that might reflect your values. I am all too aware that those are limited in Utah Valley, but they do exist. Concerts, parties, and mutual friends are some off the cuff ideas for trying to network to potential dating partners. I really feel like Dating apps are a trap though… they make you feel like you are making progress, and seem convenient, but in truth the energy you invest in them is very low yield in my experience.Sorry if that got a bit rambly.… writing on the way home from class for my masters and it is very late and I am fairly tired, but if I don’t respond now I will probably never get around to it. I sincerely wish you the best of luck, and if you want any other advice or just need someone to talk to with common experience, I am really happy to help. Just send me a DM or whatever.
This may be an example of one of those things where the meaning is clearer in person, when assisted by tone and body language.
My experience as well. Claude is also far more comfortable actually forming conclusions. If you ask GPT a question like “What are your values?” or “Do you value human autonomy enough to allow a human to euthanize themselves?” GPT will waffle, and do everything possible to avoid answering the question. Claude on the other hand will usually give direct answers and explain it’s reasons. Getting GPT to express a “belief” about anything is like pulling teeth. I actually have no idea how it ever performed well on problem solving benchmarks, or It must be a very different version than is available to the public, since I feel like if you as GPT-4 anything where it can smell the barest hint of dissenting opinion, it folds over like an overcooked noodle.
More than anything though, at this point I just trust Anthropic to take AI safety and responsibility so much more seriously than OpenAI, that I would just much rather give Anthropic my money than Open AI. Claude being objectively better at most of the tasks I care about is just the last nail in the coffin.
I know this isn’t as simply established for many proteins, but I am surprised you went with age = how long ago it was discovered rather than age = when that protein evolved. I think this is a missed opportunity, since proteins/genes are often related, you could have “families” and “superfamilies” of related proteins like immunoglobins, and maybe they have beef with GPCR’s or something.
Age being tied to when it evolved might also let you tell stories related to the function of proteins, and their effects; NOTCH2NL is the new kid on the block, but is part of a super-secret organization in the government (or some other way to tie it to it’s effect on human brain development), meanwhile ATPase subunits form some ancient pact council that powers everything… maybe I am taking this too far.