9,391 karma · joined February 21, 2007
You really see this in spades with Uber/Lyft though. I've often compared Uber prices with friends of mine and there can be a big difference for the same trip at the same time with two different people.
This phenomenon unfortunately makes me much less bullish on the AV future, because robotaxi providers will try to do this to extract the maximum cash out of users. Someone who decides to sell their car and just take automated rides is going to get gouged because the algorithm is going to figure out that they have no alternative.
One thing I've found with a lot of adversarial algorithms is that you have to demonstrate to the algorithm that you have an alternative. For example, if you consistently check prices on the Uber and Lyft apps and pick the lowest cost one, they will notice when you price out rides and then don't take them, and you will get better prices.
Similarly, with feeds, I have been finding that Facebook and YouTube want to push me short video content, even though I don't like it and don't watch it. You can push the button that says "Show Less" or "Not Interested", but they generally don't care. What works is when they show you something you don't like is to close the app and do something else. That is a signal that they respect.
U.S. law though is highly non-autistic and what you were trying to do is just as important as what you superficially did. Hell there could have been a third set of four numbers that were the nuclear launch codes. It’s not the fact that it was four numbers, it’s what you were trying to make happen when you typed them. Now of course whether they can prove what your intent was when you typed them is another matter, but generally a duress pin should be for when robbers are breaking into your house, and the government will be on your side, and not when the government will be against you.
When carpenters moved from mostly hand tools to mostly power tools, it didn't take away much of the need to understand the techniques of woodworking, it just made it all faster.
Reading a nonfiction book is very different than reading a book summary, even if (as is often the case), the ideas in the book could have been condensed way down. The point is that actually reading the whole book means you spend far more time engaging with the ideas, and so they become more deeply embedded in your neural network. In things like sports, this is just obvious, in that no one expects someone to become a good soccer player by just reading the rulebook, and obviously more practice is the main way you get better.
Starlink is clearly the best right now but there are competitors. In particular, these competitors may not have the capacity to service millions of households at $100/mo, but they definitely do for private planes at $10k/mo, especially when the service is only being used for maybe 40hr/mo.
I don't know where they're at in terms of civil engines, but each new generation of engines eeks out less and less additional performance. If China comes out with something like the CFM LEAP at a good price, I'd imagine they could sell that for many years to come.
This is the fundamental problem. You have to look at the equilibrium. When you submit a PR, you're asking for some of my time. I have to figure out if it's likely to be worth it for me. If you have a track record of producing useful software that I have merged before, you're putting your reputation at risk when you submit a new PR, so it's probably good. If you start sending AI slop, I'm going to downgrade your reputation.
If you have no track record though, I'll probably at least take a glance since even if I'm not sure, at least you had to spend some time to write the code and put together the PR. Now that's not true.
My guess is we're going to have to create some new systems for reputation, maybe bond posting, maybe "sponsored" PRs, where someone trusted vouches for it, etc.
Incidentally, this doesn't just apply to PRs. It's emails, all kinds of other messages, reports, etc.
I’d just add something else I’ve noticed with social organizations, that people used to run them more often, and one form of compensation they got was status in the community for doing something good for everyone, and that status feels like it has diminished since I was a kid. As America has changed demographically, some cultural traditions like volunteerism haven’t diffused as well. There is a tendency when two status systems live side by side, that the lingua franca is always money, and so people focus on that because having money is recognized by everyone.
It would be awesome if we emphasized this more in schools. One place to start would be talking more about my favorite founding father, Ben Franklin, who started the first public library and the first fire department in the U.S.
It used to be that almost every river was unsafe to swim in.
It’s possible to fix environmental problems. It just requires the will.
Ultimately what it means to be a professional is that you are responsible for your work. That’s why you get a salary instead of being paid by the token.
Just on a personal level I’ve found it hard for example to get YouTube and Facebook to stop showing me short videos, which I don’t want to see. You can click the “not interested” or “show less” button, and it doesn’t do much.
What works though if you apply it consistently is, when you see something on a feed that you don’t want anymore, is to immediately close the app and don’t come back for a while at least. That’s the strongest signal you can send to their recommender.
Some of these ideas strongly carry over to the idea of AIs acting as autonomous agents as well.
Given that it’s just barely possible, you can’t just make things twice as strong as you think you’d need to, just in case something unexpected happens. Anyhow when something moderately unexpected happens, that means you may get a giant fireball like we saw today.
Effectively it’s burning all of the trust built up with consumers as firewood by tricking them into buying mediocre products at high prices.
In the WHO recommendations, they say to get 75 minutes of vigorous or 150 of moderate per week. I believe in this study they use the same double counting of vigorous minutes.
I’ve seen other studies that say you get most all of the cardio benefit you can with about 150m vigorous/300m moderate. You could roughly get that by running about 2.5 miles per day.
The real question is what the purpose of prediction markets are. For sports, there isn't really much of a purpose to the markets except entertainment for the bettors, and harvesting cash for the bookies. There are also various advantage bettors (who may be involved in corruption or not), who attempt to harvest cash from some combination of the bookies and the bettors. Generally IMO these are bad due to simple human frailty though. We figured out a long time ago that for the most part, making gambling available to the general public was a net negative to society, because it mostly transfers money from addicts to big corporations, destroying lives.
For major world events, one purpose of prediction markets is just to generate a price. It's potentially useful for people to know that, in an adversarial market, what the aggregate probability of an event is. It can also be useful theoretically to hedge risks. Whether it's practical to do that depends on the depth of the market though, and with the current markets, it's not. Even more traditional "prediction" markets like commodity futures aren't deep enough to usefully hedge most risks. For example, you might think that major oil companies might hedge future pricing risks, e.g. they want to drill an oil field with a high production cost, but they're worried that the price might go down before they finish production on the wells. Generally though, the markets aren't deep enough for them to be able to do this, so they just won't drill fields that have a production cost more than roughly the lowest price in the past 20-30 years, depending on the age of the executives in charge of the decision.
There's this other purpose of prediction markets though, which is money/information laundering. People may have secret information where their employer has a strong interest in it remaining private, however the person with the information isn't that well compensated, so they monetize the information on prediction markets. On the darker side, you can have wildly illegal markets like "assassination futures" where people bet on when someone will die, and you can bet on a particular outcome, and then make it come true. There are lots of markets somewhere in the middle where someone can take an action in the context of them being a trusted agent of an organization, but instead of following their duty as an agent, they do what is profitable based on their bets in the prediction market.
Overall, IMO there are some good uses for prediction markets that allow people to hedge risk on both sides and enable useful economic activity, but most of the uses I've seen in practice are a net negative.
My impression with no inside knowledge, but understanding what Elon companies are like, is that he was assigned essentially an impossible task at Tesla and tried his very best, but it could not be done, and he semi-burned out. It makes sense for him to be getting back on the horse now.
The Elon approach to management as I see it is to assign what normally would be totally unreasonable goals to a small group of extremely bright people, and they work their asses off and somehow find a way. Sometimes this works, and sometimes it doesn't. If it works and the impossible was in fact, just barely possible, you dominate the market, everyone gets rich, and the people see it as the most exciting, intense, and rewarding part of their career. If it doesn't, they get depressed, divorced, and looking for other work. The Elon magic is threading the needle closely enough that a lot of the seemingly impossible things are in fact possible with enough hard work and brainpower, but although Elon is extremely good at this, the nature of the thing is that you can't predict which side you'll wind up on fully accurately.