I am Andrew Hyer, currently living in New Jersey and working in New York (in the finance industry).
aphyer
The board had a choice.
If Ilya was willing to cooperate, the board could fire Altman, with the Thanksgiving break available to aid the transition, and hope for the best.
Alternatively, the board could choose once again not to fire Altman, watch as Altman finished taking control of OpenAI and turned it into a personal empire, and hope this turns out well for the world.
They chose to pull the trigger.
I...really do not see how these were the only choices? Like, yes, ultimately my boss’s power over me stems from his ability to fire me. But it would be very strange to say ‘my boss and I disagreed, and he had to choose between firing me on the spot or letting me do whatever I wanted with no repercussions’?
Here are some things I can imagine the board doing. I don’t know if some of these are things they wouldn’t have had the power to do, or wouldn’t have helped, but:
Consolidating the board/attempting to reduce Altman’s control over it. If Sam could try to get Helen removed from the board (even though he controlled only 2⁄6 directors?) could the 4-2 majority of other directors not do anything other than ‘fire Sam as CEO’?
Remove Sam from the board but leave him as CEO?
Remove Greg from the board?
Appoint some additional directors aligned with the board?
Change the board’s charter to have more members appointed in different ways?
Publicly stating ‘We stand behind Helen and think she has raised legitimate concerns about safety that OpenAI is currently not handling well. We ask the CEO to provide a roadmap by which OpenAI will catch up to Anthropic in safety by 2025.’
Publicly stating ‘We object to OpenAI’s current commercial structure, in which the CEO is paid more for rushing ahead but not paid more for safety. We ask the CEO to restructure his compensation arrangements so that they do not incentivize danger.’
I am not a corporate politics expert! I am a programmer who hates people! But it seems to me that there are things you can do with a 4-2 board majority and the right to fire the CEO in between ‘fire the CEO on the spot with no explanation given’ and ‘fire every board member who disagrees with the CEO and do whatever he wants’. It...sort of...sounds like you imagine that in two weeks’ time Sam would have found a basilisk hack to mind-control the rest of the board and force them to do whatever he wanted? I do not see how that is the case? If a majority of the board is willing to fire him on the spot, it really doesn’t seem that he’s about to take over if not immediately fired.
One major point I think is under-discussed is what this means for nuclear proliferation.
Ukraine used to possess nuclear weapons, but agreed to give them up in exchange for promises of protection from the US and Russia.
With this plus the fall of Gaddafi in Libya a while back, it’s hard to see a result that isn’t ‘everyone wants to get nuclear weapons ASAP, and no-one wants to give them up.’ If promises of protection in exchange for nonproliferation aren’t upheld, there’s very little incentive for nonproliferation.
Many of these are rather hard to read. The following are my even-more-slimmed down summaries based on skimming just those few paragraphs. Some may be in error, if you disagree with any let me know:
1. Argues in favor of more bioenhancement and against refusing to do bioenhancement for reasons of ‘egalitarianism’, though in a somewhat wishy-washy way ‘we endorse a cautious proposal’.
2. Not really related to bioethics—argues in favor of pseudonymous publishing.
3. Argues in favor of ‘nudging’ patients to obtain consent for treatments.
4. Argues that if you favor assisted suicide when doctors do it, you should also favor it when for-profit entities do it.
5. Argues that all newborns should be screened for critical congenital heart diseases, without exemption for parental religious beliefs.
6. Tells the story of a patient who demanded his doctors do something really stupid and refused to budge. Unclear what moral, if any, they want to draw.
7. A weirdly meta paper that evaluates methods for evaluating ethics. I have no idea what this means.
8. Frets about off-label use of ketamine being worrying, recommends formal research into it combined with stricter regulation of off-label uses.
9. I have no idea what this paper is about. It tells the story of a kid called James, but I don’t know what it wants to draw from it.
10. Argues against ‘nudging’ people to register as organ donors.
11. Argues that traditional consent is inadequate for a ‘biobank’ (which stores biological samples of people?) and that you should use something called a ‘Participation Pact’ instead.
12. Argues that bioethicists should pay attention to neuroscience?
13. Agrees with someone called Lee that bioethics and environmental ethics should be linked, but disagrees that his idea of ‘public health ethics’ is a good way of doing it.
14. Asks about how to allocate limited supply of a drug called nusinersen for spinal muscular atrophy patients.
15. Argues that consent to organ donation is ‘vague’ in harmful ways.
16. Talks about ‘right-to-try’ trials where terminally ill patients try untested drugs. Some waffling, I’m unclear if they approve or not.
17. Argues that diversity in researchers will help make research better.
18. Talks about what to do if your patient’s relatives try to fix him with prayer and magic. Appears to argue that you should support the relatives as well as the patient.
19. Argues that conscientious objection to abortion should not keep you from being sued if your refusal causes injury.
20. Proposes a different standard for how to evaluate parent’s decisions re. medical care for their children. Unclear how it differs.
21. Argues that standards for resolving e.g. disagreement between doctors and family members on what to do with an incapacitated patient should be simpler.
22. Argues that all children should get measles vaccination. Claims that this paper’s conclusions ‘reframe the dialog on measles vaccination...from a framework of what is owed to parents...[to] the framework of what is owed to children’. Unclear why that wasn’t the framework we were in already.
23. References Gattaca, then argues (against some other bioethicist) that bioenhancement can be good if regulated.
24. Argues that ‘professional guinea pigs’ who participate in trials for money are being exploited by capitalism.
25. Argues that palliative care for children is currently in a very bad state, and that we should do research into it and make it better.
26. I have no idea what this paper is about. I don’t think the authors do either. (EDIT: second opinion is that I’m being uncharitable here, see Kaj below).
27. Declares that a 2010 study on premature babies called SUPPORT was unethical, arguing that it placed one of its treatment groups at higher risk than they would have been without the study.
28. Talks about how to handle decision-making for unrepresented patients. Unclear what they think you should do.
29. Argues that the HEC-C program (a healthcare certification program of some kind) is good.
30. Argues that the HEC-C program (again!) should be more diverse—what they seem to mean by this is not the standard concept of ‘diversity’ but that it should cover a more diverse set of medical situations.
31. Argues that informed consent is ‘nonsense’ because people sometimes believe multiple different things.
32. Argues (against some other bioethicists) that you do not need consent from a patient to test whether they are brain-dead.
33. Book review of a book arguing that medical ethics ‘cannot be regarded as an extension of common morality’ and must be treated as a completely different thing.
Sometimes other people are going to care about different regularities from what you care about.
If I am in a botany laboratory, and a botanist instructs me to ‘put the fruit samples in Refrigerator 1 and the vegetable samples in Refrigerator 2’, I will...well, I’ll ask for clarification first, but if I can’t do that I will put tomatoes in with the fruit samples. This is because botanists care about the regularities for which tomatoes are similar to other fruits (like being made of flesh and full of seeds).
If I am in a kitchen, and a cook instructs me to ‘put the vegetables in the salad bowl and the fruits in the fruit salad bowl’, I will put the tomatoes in with the vegetables. This is because cooks care about the regularities for which tomatoes are similar to other vegetables (like being tasty in a salad and not in a dessert).
Both of these definitions of ‘fruit’ are valid, depending on context.
If you show up in a kitchen and demand that the cooks stop putting tomatoes in salads, soups and savory sauces (like you do with vegetables) and instead put tomatoes in the fruit salad, make a tomato crumble, and a tomato ice cream sundae (like you do with fruits), because according to botany tomatoes are fruits, then you are a lunatic.
If you show up in a kitchen and demand that the cooks say ‘put the vegetables plus the tomatoes in the salad bowl, and the fruits except for the tomatoes in the fruit salad bowl’, you’re at least not screwing up dinner, but you’re being annoying by denying cooks access to a useful regularity that they care about.
Numerical cost-benefit analysis.
(Disclosures: am American. Strong views presented without evidence.)
The most damning indictment of French food I can think of is the fact that American capitalism hasn’t even bothered stealing it. We have Italian restaurants on every corner, Chinese and Mexican and Thai and Indian and Korean and every other cuisine from every other corner of the world...except French. One time I went to a Korean hotpot place, but it was too full and had a long wait, so instead of waiting I walked to a different Korean hotpot restaurant. There are two German restaurants within a few minutes’ walk of my apartment, because even if German cuisine isn’t a contender for the top spot it’s at least good at what it does.
And then we have one very fancy French place per state that stays in business by pandering to a tiny number of pretentious food critics who care about every aspect of food except its taste.
“Good point”, says Irene. She opens the second box, and goes home with $1,001,000. Why shouldn’t she? Omega’s dead.
SIMULATION COMPLETE.
RESULT: BOTH BOXES TAKEN.
OMEGA-COMMITMENT CONSISTENT ACTION: OPAQUE BOX EMPTY
Irene promptly walks up to the opaque box and opens it, revealing nothing. She stares in shock.
There’s a model of white-collar employment I think is missing here. (I also thought it was missing at some points in the Moral Mazes sequence, but never got around to writing it down then).
The model is underutilized employees as option value.
Imagine yourself as a manager, running a small team at some company somewhere.
Most weeks, your team has 40 hours of work to do.
Every so often, there is a crisis. Perhaps your firm’s product is scheduled to release in 2 weeks when a regulator suddenly dumps 500 pages of compliance questions on you and refuses to approve your product until they are answered. Perhaps a security flaw is discovered in a major library you use and you need to refactor your whole codebase. Perhaps someone makes a mistake and your firm’s biggest client is ticked off. In any case, you have 200 hours of work to do that week, and if it is not done that will be Very Bad.
How many employees should you hire?You could hire one employee. This employee would be busy but not unmanageably so most weeks...and then as soon as something went wrong, there would be a disaster.
So instead you hire perhaps four employees. Most weeks they each have 10 hours of work to do, and spend the rest of the time chatting around the water cooler or playing solitaire on work computers or whatever. And when a fire drill happens, you get them to drop the solitaire games and maybe put in a few extra hours and you have enough people to handle it.
The thing you are buying with these three extra employees is not the 10 hours they each work in a typical week. It is them being around and available and familiar with the job when something goes wrong.
-------END OF MODEL, BEGINNING OF ARGUMENT------
Many people seem to me to be operating on a model something like this:
“There is X hours of work a week to do in your job. If you do those X hours of work, you are Doing Your Job. Employers want you to be e.g. in the office 9-to-5, even if that is not needed to Do Your Job, because they are evil monsters who enjoy the taste of your tears.”Under that model, if you can do your job in 10 hours a week with GPT that is great! This high productivity should lead to some improvements, either in you having more free time, or in you being able to get more jobs and make more money.
Under my model, of course you can do your job in 10 hours most weeks. This has nothing to do with GPT! Your job can be done in 10 hours most weeks because most weeks nothing is on fire.
But if you have two employees:
Alice shows up at the office, does 10 hours of work, and plays solitaire for 30 hours.
Bob works from home, does 10 hours of work, and then switches to Job #2.
Bob is actually worth much less as an employee. Because when something goes wrong, you can grab Alice, call in your option, get her to stop playing solitaire, and have her do a bunch of work that was needed...but when you try to call your option on Bob, he’ll be on a call with his manager from Job #3. Your actual binding constraint is how much work needs to get done during a crisis, and Bob contributes very little to that.
My client acknowledges his guilt but wishes to appeal the harshness of the punishment.
My client has renounced his evil ways, and sworn to follow the straight and narrow path. As proof of his sincere remorse, he has submitted two new articles (1, 2), intending to pursue Good Hearts with honestly-written articles earning honest upvotes.
However, the severity of the fine has made it impossible for him to ever return to a normal life in society. The GH1000 penalty is as of this writing ~5 times the largest quantity of GH earned by any writer.
My client is additionally willing to delete the criminal posts, just as soon as someone tells him how to delete a draft (it’s unclear to me how you can actually do this).
In the interests of clemency and rehabilitation, Your Honor, I ask that you reduce my client’s penalty to a sum he can more realistically ‘work off’ through good deeds, rather than fining him an amount far in excess of what even the wealthiest GH-holders are able to pay.
We don’t care about how many FLOPs something has. We care about how fast it can actually solve things.
As far as I know, in every case where we’ve successfully gotten AI to do a task at all, AI has done that task far far faster than humans. When we had computers that could do arithmetic but nothing else, they were still much faster at arithmetic than humans. Whatever your view on the quality of recent AI-generated text or art, it’s clear that AI is producing it much much faster than human writers or artists can produce text/art.
I have learned that I do not want to go on an adventure.
Some of the feedback was disappointing– people said things along the lines of: “Lies of omission are not lies! The fact that you have to add ‘of omission’ shows this!”.
The worst that I have gotten in that vein was: “When you are on stand, you take an oath that says that you will convey the truth, the whole truth, and nothing but the truth. The fact that you have to add ‘the whole truth’ shows that lies of omission are not lies.”
This was from a person that also straightforwardly stated “I plan to keep continuing to not state my most relevant opinions in public”, as they were defending this behavior.
What are your...let’s say 5...most controversial political viewpoints that you would get in most trouble (with family/friend groups/employers/government/etc.) for expressing?
You might prefer not to answer this question. Tough luck! Lies of omission are lies too, and you shouldn’t say them.
You might prefer to have controversial conversations in private with people you trust, away from Twitter frenzies and the pressure to dunk on opponents. Tough luck! You see,
the public cost (such as being dunked on) is the correct update from the public
about you holding these positions, and
Without this public ritual, [you] can hold [your] opinion, free of charge.
This may not be what you intended to say in this piece. But it does appear to be what you actually said.
I had a similar gut reaction. When I tried to run down my brain’s root causes of the view, this is what it came out as:
There are two kinds of problem you can encounter in politics. One type is where many people disagree with you on an issue. The other type is where almost everyone agrees with you on the issue, but most people are not paying attention to it.
Protests as a strategy are valuable in the second case, but worthless or counterproductive in the first case.
If you are being knifed in an alleyway, your best strategy is to make as much noise as possible. Your goal is to attract people to help you. You don’t really need to worry that your yelling might annoy people. There isn’t a meaningful risk that people will come by, see the situation, and then decide that they want to side with the guy knifing you. If lots of people start looking at the situation, you win. Your loss condition is ‘no-one pays attention’, not ‘people pay attention but take the opposite side’.
And if you are in an isomorphic sort of political situation, where someone is doing something that basically everyone agrees is genuinely outrageous but nobody is really paying attention to, protests are a valuable strategy. They will annoy people, but they will draw attention to this issue where you are uncontroversially right in a way that people will immediately notice.
But if you are in an argument where substantial numbers of people disagree with you, protests are a much less enticing strategy, and one that often seems to boil down to saying ‘lots of people disagree with me, but I’m louder and more annoying than them, so you should listen to me’.
And ‘AI development is a major danger’ is very much the ‘disagreement’ kind of issue at the moment. There is not broad social consensus that AI development is dangerous such that ‘get lots of people to look at what Meta is doing’ will lead to good outcomes.
I have no actual expertise in politics and don’t actually know this to be true, but it seems to be what my subconscious thinks on this issue.
Visits to emergency rooms might not be down if parents are e.g. panicking and bringing a child to the ER with a bruise.
If I think that this post is interesting and well-written but disagree with it and prefer the current karma system, should I upvote or downvote it?
I am a human, living on a ball of rock third out from a star in a galaxy I call ‘Milky Way’… This is an additional group of words, group two, out of my quick try at writing what you ask for… Yakka foob mog grub puggawup zink wattoom gazork chumbul spuzz tirv yo klii voy tryvial toy nikto savaran to dillino...
It is hard to avoid a symbol that is common in most writing, but so far I am avoiding it...
Although following such constraints is annoying, I can accomplish such a task, which GPT cannot grasp at all… Svirak nilliak toynodil r saasak motooil bordonak xi toy vir nolonol darak, silvi sansak so, tirin vak go na kilian ay BOY no vak...
If artificial minds cannot grasp constraints such as this, it is hard to claim that such minds truly ‘grok’ what is output… By contrast, I find this a fairly straightforward task, though slightly annoying, and will find a bit of joy in finishing it shortly...
Aphyer
The ‘full repayment’ part is only sort of true, in a similar way to what happened with MTGOX, due to bankruptcy claims being USD-denominated.
Suppose that:
You owe customers 100 Bitcoin and $1M.
You have only half of that, 50 Bitcoins and $500k.
The current price of Bitcoin is $20k.
You are clearly insolvent. You will enter bankruptcy, and the bankruptcy estate will say ‘you have $3M in liabilities’, since you owe $1M in cash and $2M in bitcoin.
Suppose that the price of bitcoin then recovers to $50k. You now have $3M in assets, since you have $500k in cash and $2.5M in bitcoin! You can ‘fully repay’ everyone! Hooray!
Of course, anyone who held a Bitcoin with you is getting back much less than a bitcoin in value, but since the bankruptcy court is evaluating your claims as USD liabilities you don’t need to care about that.
This ‘full repayment’ is plausibly still important from a legal or a PR perspective, but e.g. this part:
there is typically a legal fight over whether a company was insolvent at the time of the investment or that the investment led to insolvency. If every FTX creditor stands to get 100 cents on the dollar, the clawback cases that don’t involve fraud wouldn’t serve much of a financial purpose and may be more difficult to argue, some lawyers say
is better thought of as ‘our legal system may get confused by exchange rates and pretend FTX was always solvent’ rather than as ‘FTX was actually always solvent’.
I agree with this in most cases. I do think there’s a distinction to be drawn between ‘user tricks the model into saying racist word’ and the ‘model aggressively gaslights user’ dialogues that have been floating around from Bing Chat—the latter seem at least closer to an alignment failure.
Quantum turnip million. RELEASE malarial assemble!
FTFY.
While I have no knowledge of or views on the situation above, this is just a good thing to do in general? Like, most sentences that begin with the phrase ‘my boss, whose house I live at and who I have only a handshake agreement with on pay...’ are not going to end well.