• 3 Posts
  • 451 Comments
Joined 2 years ago
cake
Cake day: June 5th, 2023

help-circle
  • I think that the general idea of artificial intelligence in education hold some promise, in the sense that if you could construct a machine that can do much of the work of a teacher, it should enable kids to be taught in an individual way currently only possible for those rich enough to afford a private tutor, and such a machine would be labeled as an AI of some kind. The trouble is, like with so many other things AI, that our AI technology just doesn’t seem to be up to the task, and probably just won’t be without some new approach. We have AI just smart enough for people to try to do all the things that one could use an AI for, but not smart enough for the AI to actually do the job well.



  • Nuclear rocket engines. A bit less ambitious than most of the responses, but most things here seem to either refer to technologies we don’t have yet but seem within a century or so of developing, which doesn’t fit the question, or vague consequences that one wants that tech to have without it being clear how our current technology gets there. But nuclear rockets definitely fit the question, because we have built and ground tested them before, decades ago even, we just haven’t bothered to actually use the things. And they should theoretically make developing things like space industry or manned space exploration beyond the moon more viable, by being more efficient than chemical rockets while giving better thrust than ion engines do. They don’t work well for launching from the ground, but since our launch abilities have increased a fair bit in the past decade or so, actually getting the things to space in order to use them should be easier than ever.




  • That, and the way that the text boxes have a sort of old yellowing newspaper tint to them instead of just being white. Not sure why AI does that, I’ve not really used it but it feels like back when it couldn’t do text very well, AI stuff people posted at least didn’t have the sort of washed out sepia colors it does now whenever someone asks it for a cartoon art style.










  • I dont know the economic stats on what percentage of companies have unions, but theyre not exactly non-existent, I know people that work unionized jobs, a place I used to work for had one (not that I saw it do much, but I wasnt there that long), and the business I work for has them for some of the countries it operates in (mainly ones in Europe I think). They might not exactly be the norm in the US right now, but they’re not some fantasy either. And I would imagine most companies with one have the resources to deploy something like this if they have a use case where it would actually make any sense to. Maybe not train a leading AI model from scratch given the expense numbers I keep seeing reported on that, but that doesnt sound like what this kind of application requires.


  • I know that. My point wasn’t that automation will make companies behave differently, but that the maximum demand that can be forced upon a business by things like unions is increased if the pool of money they can demand from before the business can’t operate anymore is larger. What I said is applicable for economic systems beyond capitalism, for that matter, since it’s just a more specific way of saying that the average person can theoretically have more things when the average number of things made per person increases.


  • You misunderstand my point then. There are ways to force a corporation to pay people more (unionization, minimum wage laws, sufficiently bad labor shortages etc). There is a maximum amount of wage that these things can extract out of a company, because if the labor costs grow enough to make a business unprofitable and they’re unable to either raise prices or cut things enough to compensate, then that business will shut down instead. Increasing the amount of revenue per employee raises this theoretical ceiling on what can be paid. The method to actually get them to pay that wage is beyond the scope of my point, just that whatever method one might prefer has a higher maximum on what it can get when productivity is higher.