Yeah, generative AI used to be wild, alien creativity and not something that made art kids furious.
I wonder if models can be trained for "high-temperature" purposes. I'd rather have a model which can surprise me than one which can predicably produce generic mediocre results. I mean you can run them on high temperature of course, but it doesn't seem like it's optimized for that.
Passive revenue streams are dependent on someone else not having access to those passive revenue streams.
Yes, it would be great to be free of debt, but for me it would have to mean moving away to somewhere real estate prices are not only low, but dropping for all too understandable reasons. And also a huge distance away from friends and family. There's a reason people mostly don't do this, and it's not that they feel a moral obligation to corporate.
>Passive revenue streams are dependent on someone else not having access to those passive revenue streams.
I think many assume it is some sort of zero-sum-game. This simply isn't how most unique successful product and service based niche businesses operate.
Most firms that directly try to hyper-scale their way into market dominance simply fail within a few years. The smarter bros often cash-out after the IPO these days.
Some people do feel entitled to others free time, and post-unknown-risk capital investments. Those folks can't help anyone succeed at anything except bankruptcy. =3
Maybe it feels that way, but that's an illusion, then. The things are often haunted by poor decisions taken when they were built and take a ton of maintenance. Concrete isn't as maintenance-free as it seems!
I think that's narrowing it too much. Edward Hopper's "Sunday Morning" evokes the same feeling with me - the feeling of something that just was there, that we passed through without paying much attention to it at the time, but which we now remember.
Maybe you wouldn't have that feeling about a beautiful old European town, but I bet some of the people living there would.
I don't think "no people" should be seen as a hard rule. The Hopper paintings mentioned in the article, for instance, usually have a few people in them, and yet he's clearly aiming for the same thing as modern liminal spaces pictures and achieving it. Hopper paintings have only become more unsettling with time, because they feel so modern, like that was yesterday, when you know that objectively it was over 100 years ago.
I don't believe blackmail is the most effective thing. The problem with blackmail is that even if you do as the blackmailer says, they still have the threat hanging over you forever. That increases the risk that they do something desperate.
What is effective, isn't blackmail, but complicity. Doing bad things together. Then you get a shared interest. The people Epstein had blackmail on, knew that he wouldn't use it casually, because after all there would be no way to use it without implicating himself. But if he were desperate, he might. So the victims had an interest in keeping Epstein not desperate.
So it's the bad things they do together which is dangerous. Even things they do in full view of the public can work, because the threat isn't necessarily that the public finds out, it's that if one is held accountable, then all are at risk.
I think it's closer to dragging them into a conversation. Something along the lines of: "Hey I know you did X, let's talk about Y and maybe to Z since you're already in deep shit."
I think what made Epstein effective was balancing the blackmail with favors for complying. If all you do is get dirt on people, eventually it will fail. But if you give them something too, they are less resentful.
Good question. I think maybe it was because a good deal of academics from the Islamic world were educated in France already due to the colonies they did have. Algerians, Tunisians, Syrians, Moroccans, Lebanese... Not an expert so could totally be wrong here, but maybe France put in more effort to incorporate local elites into their own bureaucratic elite system? I imagine there were probably more Muslims in the grande ecoles in France in the 50-60s than in the English elite schools (or US ones for that matter).
Also, I know France practiced (and probably still practices) a good deal of academic "cultural outreach", promoting their culture and especially their language abroad.
Well, we've all been students, haven't we? And most of us probably have experience with ways of teaching us that worked, and ways that didn't. Of course we're all going to have an opinion.
I don't have any grand theory of education, but I have some stories of what worked for me and what didn't.
I learned English from a guy with a radical method: the "direct method" or "natural method". After the first lesson explaining what he was going to do, he spoke only English in class. The textbook also had only English (vocabulary was taught with pictures). This was about third grade elementary school. This worked great for me, I always had top marks in English. German, by comparison, was always taught to me in the traditional method with grammar lists etc. durchfürgegenohneum, ausbeimitnachseitvonzu, and I still remember that crap and I still absolutely suck at German.
So one "revolutionary", running his own radical program (he would never have been allowed to do that today), helped me. I think we should let people try things.
I'd agree with this conclusion from another angle as well. It seems slightly odd to me that people think there must be a single "right" way to teach. What works for one student, one group of people, doesn't necessarily work well for another.
And it also goes the other way as well. One form of pedagogy might work excellently for one teacher, yet he may do abysmally at another. What's "right" for him may be wrong for another teacher. By striving for something like homogeneity you disadvantage not only students, but also teachers.
This is all even more true in current times as educational outcomes continue to decline even as ever more money is pumped into education, and teacher churn rates are at record highs, with many completely leaving the profession.
Humans are not so different from one other that we need different ways. However there are a lot of ways that work and it is very hard to run a real study to figure out which is best. You cannot isolate all the variables (several of the different ways claim teacher quality is important - just one variable that is hard to isolate)
Why do you think? For an example of something in support of my argument, China (and a number of other East Asian countries) use a very hardcore memorization + training routine. And they are having literally the best educational outcomes in the world from it, but such a thing would almost certainly fail catastrophically in a contemporary American classroom.
> such a thing would almost certainly fail catastrophically in a contemporary American classroom.
It definitely would fail but isn't it an order of magnitude more likely that's due to the parents, teachers' unions, and other factors rather than American students are neurologically different than Chinese students and therefore learn differently?
If they do have much better outcomes (I have no idea if this is the case or not), if you made that change in kindergarten today and moved it up through the end of high school with that class, I bet you'd see remarkable improvement in them compared to older cohorts.
Yeah, here [1] are the PISA outcomes. PISA is an international test that's generally the gold standard for comparison of educational outcomes on an international level. [1] Over the last testing year China was #2, the test prior #1. Singapore was #1 in the most recent period, and is around 75% ethnic Chinese.
Whether the differences are genetic or cultural is interesting but doesn't really matter. The reality is that they exist and are relatively immutable. For a very basic example, in China failing students fail. In American schools, failing students tend to be passed along. And such things are difficult to change, even if you could prove beyond any doubt that doing so would yield better outcomes for everybody.
The real question is how you 'do at life'. Some tests are a proxy but not all. And even when it is, you can 'teach the test' and get worse results at life even though the test itself is good. Too often we are not sure what is a good proxy.
That is the grain of salt all education comparisons need to be made with.
They went from literally starving to death in the 60s to becoming one of the greatest, if not the greatest, power on Earth with a thriving and growing middle class - which is really the backbone of a nation. So I think it's safe to say that their life outcomes are on parity with their measurable educational outcomes.
It's hard to say because the US was, by far, the most dominant economic power in the world in decades past, and now we're losing that role, or already have depending on how you measure things. But the nature of going from #1 to #1 (or #2 as it may be) means you have no real basis for comparison how things might have been.
But what I'd observe is that by many metrics the US hasn't really made major progress since our glory days. For instance real median wages from 1979 to present have increased by less than 13%. [1] It'd be disingenuous to compare that to China because they started from much less so obviously you expect dramatically higher growth, but I think it's reasonable to claim that we could have, and should have, done dramatically better than 13% growth in almost 50 years.
Another point is that the US was able to 'artificially' strengthen its economy owing to the dominance of the dollar and being the center of the tech boom. I say artificially because those things were obviously going to be liminal. We needed to leverage those advantages into something long-term. Instead it looks like we've been mostly treading water and largely wasted it all.
And now pair all of this with contemporary times with a sharply divided population, unsustainably low fertility rates, terrible education outcomes, and more. And no, I don't think it's reasonable to claim that the US is doing well, let alone to the point of being able to rest on our laurels as you're implying.
While China's improvements are amazing, don't be fooled, overall China is still behind. They have more people and so totals can look good, but the common person has much improvement left before they beat the US. Ultimately, I think your argument comes down to, when you're on top, it is very hard to improve much more, while, when you're starting at the bottom, you can improve a lot very quickly and easily. Indeed, China's growth numbers have been slowing a bit recently, which supports my thesis. Only time will tell of course
Also there is plenty of reason to think that China isn't giving honest numbers, but what's the true numbers are is impossible to know
Behind by what metric? PPP their economy is already much larger, their educational outcomes are much better, they're doing wild stuff in bleeding edge domains like robotics, built/launched their own space station super rapidly, and so on. Pretty much the only domain they're clearly behind in is rocket technology owing to SpaceX, and back here stateside you have plenty of people that hope the company somehow fails because they don't agree with the political opinions of its founder.
And the numbers we're citing aren't just from China, but from third party organizations. In any case, the arguments don't come to debating numbers - the change is obvious and visible. It's not like they're claiming new record low unemployment while seemingly large numbers of people somehow people struggle to find jobs.
Are they having the best results from that? I've seen the claim of other countries using that and having book smart kids who can't think. (Whatever that means)
There is a common want to make the grass greener. However it isn't always and most people don't know.
I wonder if models can be trained for "high-temperature" purposes. I'd rather have a model which can surprise me than one which can predicably produce generic mediocre results. I mean you can run them on high temperature of course, but it doesn't seem like it's optimized for that.
reply