Physicist and dabbler in writing fantasy/science fiction.
Ben
Karma: 2,809
I think this sounds surface-level convincing but for me it seems to fall down on further analysis.
Lets say you are a mob boss. You can rob someone and make money. You can also employ underlings to rob more people. In a sense the underlings are middle men you cost you money compared to the counterfactual where you robbed just as many people single-handedly.
You are a sugar or cotton consumer is 1700′s England. You can get cheaper goods by stealing the labour to produce them. But moving to the carribean and buying wips isnt conveninet for you so you use plantation owners as middle men.
If you dont care at all about the misery of the slaves, the extra labour taken from them is kind of just added economy. Like, in some imaginary world where we could trade with chickens would we be better of doing that than employing battery farm owners as ‘parasitic middle men’. The point is, we are all parasiting on the chickens (egg and chicken consumers). Similarly, the sugar buyers of the past were part of the parasite that stole slave labour.
I had a vaguely analagous experience where an AI interviewed me for a job last week, surface level appearance of sanity (mostly) somewhat hilarious dysfunction underneath. I think companies think they can get away with it through some combination of hype and AIs getting better
Maybe if they have a call centre abroad where most operators speak very poor English, then poeple leave bad reveiws complaining about it, those reviews are still online and embarasing 2 years later.
They deploy an AI. Equally bad reviews, but this time they are only brefily embarasing because 2 years later anyone reading them thinks ‘oh, this review is 2 years old, of course the AI was bad then, but now...’
Hopefully its short lived, people like your doctor will hopefully write to the company and say something like ‘as a matter of patient safety we will no longer be accepting drug requests unless they are made by a human being. Please sign our updated terms indicating that all drug requests made to us in future are finalised by a human.’ That rules like this will be set by businesses, insurance companies and governments is kind of the obvious blowback against premature deployment.
Contrary to most comments here, I was actually quite surprised at how good the LLMs story was. It actually had a premise, a core idea, that seemed novel and prefectly serviceble that is something I haven’t seen one manage before.
I was forced to stop reading halfway, but came back to finish it. Which means that for me it passes some kind of natural test.
Also, on a meta level the most obvious joke in the world is to say the story is perfect, but complain about the perfection. So the LLM is inviting us to call its work perfect. Clever robot. Good social engineering.
I agree that their posaibly are situations where belief influences outcomes in such a way that factually incorrect beleifs can produce better outcomes. For the most part I think these situations are comtrived and rare, but accept they do exist.
However, in real life, we don’t encounter one choice ever, but many choices one after the other. In the long run its not about how good the initial descion making is, but what we learn. Something goes well, we ask what we did right. Something goes wrong, we ask what we could have done better. Learning well is the long term best strategy.
If I have somehow adopted wrong beleifs on purpose how can I ever assess if this is working or not? I preumably no longer know that those beleifs are wrong.
Ben’s Shortform
A dumb thought I had the other day “I suppose I am not totally convinced of empiricism, if I saw strong enough evidence against it I could change my mind.”
What is weird about this, is that in an eternal steady state universe the number of living observers in a given time period is much larger than the number of observers (including Boltzman brains) in a simialr slice of a universe approaching thermal equilibrium. So, depedning on how your Anthropic reasoning work, you could argue from the fact of your own existance that the universe must be steady state, regardless of any evidence otherwise.
(This is kind of a tension inherent in all Anthropic arguments, I am more likely to ‘be me’ (as opposed to someone else, like a Boltzman brain) in a universe with fewer people, but most possible observers are in the crowded universes, so I am more likely to be in a crowded one. Similar logic, threading oppsosite conclusions.)
A sort of opposite conversational thing I see which I hate is something like:
Bob : I am really sorry, I made a big mistake and (something has gone wrong).
Alice : How did this go wrong?
Bob : We had that event that day and we were all distracted talking to the visitors and eating the cake. I wasn’t used to the new UI yet, and it looks like I must have muddled some of the inputs.
Alice : Well, that is no excuse.
Bob : I never said it was. You asked for an explanation. I owned this mistake in my first line of this dialogue.
The issue here is that Bob is trying to provide a real explanation, of the kind that would be useful in providing a plan to avoid similar issues in the future. But, Alice is instead trying to give him a dresssing down for his mistakes. Possibly both conversations should happen, but not at the same time and both parties should know which one they are in.
(Although I still think asking for an explanation, and then complaining about excuses is a jerk play that is surprisingly common.)
I think leogao meant that the readers were being pretentious in their discusions of the book (looking at tea leaves).
Age of Empires—as a child, in RTS games I always massively underinvested in economy. I wanted to build all the cool units and buildings, and I kept having this idea that my economy was already huge because it was however many times bigger than it started. (In Tiberium sun I beleived having 3 harvestors was huge because in my head it was 3× default.). When playing these games with humans you suddenly realise you are thinking too small. They are just doing everything at 5x your scale and it doesnt really matter what plays you might be making right at that point. If Dominion teaches you the game is finite so you should at some point switch from building for the future and start scroing points now, RTS games teach you that exponentials are real and that if you dont spend the early game growing for the later game you should ensure there is no later game.
All games—playing the game as its creators clearly intended, vs playing to win.
I think reviewers have two seperate tasks when assessing a paper.
The first, that (in my opinion almost all reviewers are good at) is identifying the easy wins that will improve the paper. This is the land of typoes, the occssional confusing sentence or a missed citation.
The second (that reviwers are on average less good at) is telling the editor if they think the paper, at its core, is actually any good, assuming some fixes and repairs.
I think this analysis is maybe not right. The systems that endure are the systems that survive, so its less relevant in the long term how effective a system is at waging agrresive wars. Its how good it is at fighting defensive ones is the key.
The counter insurgencies in Iraq and Afghanistan were not existential for the continued survival of the USA. So they were not going to create much (or any) Darwinian pressure to select for or against societies like the USA. (Baring the, potentially important possibility of the USA leaving societies in its own image in these places).
Yes, I was thinking primarily of real life conversations with a person you already know. Where I have seen this play out as I described. Although, obviously I did not try out the counterfactual.
I have no kind of data on this, but my feeling is that it carries over to social media, at least for famous-ish people who have already had the space to build rapport. As a topical example, some right wing people in the USA (Tucker Carlson I think) seem unhappy with Trump’s recent Iran policy. My suspicion is that for MAGA types someone like this going “off script” and being against the war is significant, more so than if they encountered someone saying (for example), “Yeah, I supported Trump in his first term, but after the Capitol riots I became a Democrat.” Part of that is because the “I joined the other side over X” has to be something old, in the past, where opinion has hardened and in any case its in the past, but part of that is also the inherent implausibility of the “Side switch” bit. The previously-Trump-supporting person who turns against him after the capitol riots isn’t going to fall in love with all kinds of Democrat coded policies they previously hated.
The thing that makes this kind of threat particularly pointless, is that saying “I am on your side on all these other issues, but on this thing you are just wrong” is probably way more persuasive to people.
Lets say you are a big fan of a leftish politician like Jeremy Corbyn (JC). Someone you know to be generally on the political right tells you that they hate JC, and that in fact the awfulness of JC is WHY they are on the political right. You probably just dismiss them. But now someone comes to you who agrees with you on all kinds of policy positions, but that they tell you they dislike JC for such-and-such a reason. That seems more likely to generate a real update.
I know nothing about the underlying price strucure, but I do know that if a journey contains a sub-jounrey, that charging less for the whole than for one of its parts cannot possibly make logical sense. I had this last year when buying a baby pushchair, buying just the chair I needed would have cost me more than a package deal that included that exact chair along with some other stuff I wasnt bothered about. I went for the package deal. The package can of course reasonably cost less than the sum of its parts, but it cant cost less than any one of its parts.
Targetting the ‘break of journey’ seems weird to me. I have never done it intentionally to get a cheaper ticket, but I have, for example been on a train into Bristol and been phoned up to say that some change of plan means its eaiser to meet me a station earlier (Parkway), so I get off one station early, rather than wasting time going in a circle. Banning break of journey would sort of be like that pushchair company continuing to have the package deal cheaper than one of the products it includes, but then sending someone around to force me to use the extra stuff included.
I am curious why it is a hard problem for the prices to emable people to spend less to buy more, and, if the prices do have that quirk, why its really a problem for customers to exploit it.
I haven’t checked wikipedia for that, but I don’t think the trend is surprising. I think that a big part of it is just that it feels creepy to comment on a specific person’s attractiveness in polite company. And Wikipedia is polite company. If an actress is particularly pretty it would feel kind of weird for someone to add that to their wiki page. I can half remember a tv interview from the early 0′ies, I think it was Johnathon Ross interviewing Kiera Knightly about her new film, and he said something which implied that he thought a big part of her appeal as an actress (or even the main part) was that she was really attractive. Regardless of the truth of that position, it seemed that both she and possibly also the audience found it a really cringe topic of conversation.
I am sure there are other websites (which are Not polite company) where you will find a lot of people talking about how attractive they find one celebrity or another.
Yes, it can work. Reflecting more I think the issue is maybe that you need to be clear from the beginimg wether you are telling your overseas workers ‘your timetable will suck but we are paying super over the odds to cover that’ or are doing something more ‘normal’. (+100% vs +30%). The first was never the bargain in my example above, hence some of the frustration.
Both have advantages and disadvantages. In the former, the employees will just accept the timetable. But, you will mostly get younger, more junior (single) people, and they wont be the most capable or best qualified people, who will go for something normal. I suspect high turnover, a couple of years of highly paid nocturnal behaivior to then take that experience to try and get a more normal job where you can actually have a family, makes sense.
Instead of a ‘half move’ you have the alternative of a ‘full move’, where you move the whole operation (taking the people you need with you). On a much smaller scale some software companies did this a decade or two ago by moving from London to Bristol where property prices were lower (very short move, they are only 1.5 hours appart by train.)
I remember a Scot Alexander post a while ago about Bhudism and suffering, I beleive he was asking Isur about some aspect of it. His phrasing implied that the idea that Bhudism has something important to teach us, some kind of magic juice, was to be taken very seriously. I imagine an equivalent post about Hinduism or Islam, or even Kabalistic stuff, would have used more detached ‘they beleive this stuff’ phrasing.
I dont agree that Bhudism is somehow uniquely unhealthy to people. I do find it interesting how it seems to provoke different instinctive reactions than other religions.
I second all this.
Even when it does technically work, people underestimate the social dynamics. I knew somone who worked in Singapore for a Canadian company. Her dinner would be going cold on the table and her children wondering why mummy wasnt joining for bedtime stories while her teleconferenced meeting in Canada overrun by an hour as they all complained about having to get in at 8am and how the weather was bad in Toronto. Complaints from her and others in SG soon made it clear that these meetings had to be done as fast as possible and that mentioning anything off topic, eg the weather, random pleasantries, traffic, was liable to result in interuption/complaint from someone in singapore.
So, you can have the meetings, but freindly chit chat, going off topic or similar is completely shut down. That worked much better, shortening the meetings and letting the signapore lot at least feel their time was being valued, but doing meetings that was is not natural to people.
If they had tried some weird ‘work in a canadian time zone’ thing then the person I knew, and probably the whole team, would have quit. Maybe unmarried 20-somethings could do that for a little while, but as they gained partners and kids it would stop working for them.
I think that the worlds as it looks when we are growing up is forever set as the ‘normal’ world, and everything else is weird.
So, older people are probably more happy to think of the world as weird.
‘Normal’ and ‘likely’ are not the same. I always knew that in some sense the world we are in is unlikely but for me, the moment I really felt that I was in an unlikely world was when Trump’s ear was brushed by that bullet. Like, in the he dies or lives coin toss that is ‘it lands on the edge’ territory. And it plausibly had a huge impact on the world.