Because my netflix subscription is cancelled specifically because the "Finding something I want to watch drains my energy" phenomenon. Gradually over the course of like a year I got more and more frustrated with being suggested things, and not having a good way to find things.
So... immigration services in the US don't use criminal language when discussing how they handle people accused of immigration offenses, because there's a whole legal structure to pretending it's a civil infraction and thus you don't have any rights related to say... trial by jury or the state proving your guilt beyond a reasonable doubt.
But. The mechanical processes can include indefinite detention in facilities that look and function exactly like jails.
So... what you CALL it is almost certainly something different than what it is.
No, I think that's unfair. I, as a user, could very reasonably want a parody or knock-off of Indiana Jones. I could want the spelunky protagonist. It's hard to argue that certain prompts the author put into this could be read any other way. But why does Nintendo get a monopoly on plumbers with red hats?
The way AI is coded and trained pushes it constantly towards a bland-predictable mean, but it doesn't HAVE to be that way.
Sure, but also as an adult, you're expected to have learned some resilience to failure. You're expected to be able to be able to withstand some criticism and see negativity as a chance to improve.
I know it doesn't always work that way, but a lot of times our failures aren't just "on us", but affect others.
Children are most hurt by low expectations. Especially young children.
There's subtlety to this, high demands are not high expectations. If there's consequences for not meeting some high standard you set for children, you're going to create a very life-destroying kind of learned helplessness. Kids shouldn't be punished for failure.
And if it's something dangerous to try, then of course it's gotta be something you limit.
But beyond that, just don't assume kids aren't ready for something without evidence. Let them try.
I mostly agree, but struggle with saying this with perfect certainty.
Understanding in the "have mental model of the world, apply it, derive thoughts from that model, derive words from thoughts" pattern is a thing they don't do the way we do.
But understanding of some kinds CAN be encoded into tokens and their relationships. They're clearly capable of novel, correct inferences, that are not directly contained within their training sets.
I all-but-guarantee my "My fish suffocated when I brought it to space, even though I gave it a space suit filled with pure water, why?" test case is not something it was explicitly trained on, but it correctly inferred "Because fish need oxygenated water"
The press release writer was a bit shit at understanding basic scientific jargon. The research only finds epigenetic markers. There are no genetic changes. There are detectable changes to gene expression, and that's it.
Not really? It's a lot of work, a multi-week project, but reading a couple hundred word speech can be done in 5 minutes, following a checklist in hand, probably 10 minutes. Times 12 categories, and 80 years of history, that's a lot of time 160 hours, a working month. A lot of effort but humanely doable.
I use python repl as my primary calculator on my computer.
1. I don't have problems like the IOS problem documented here. This requires me to know the difference between an int and a float, but pythons ints have unbounded precision(except if you overflow your entire memory), so that kind of precision loss isn't a big deal.
2. History is a lot better. Being able to scroll back seems like a thing calculators ought to offer you, but they don't.
3. In the 1-in-a-hundred times I need to repeat operations on the calculator, hey, we've already got loops, this is python
4. Every math feature in the windows default calculator is available in the math library.
5. As bad as python's performance reputation is, it's not at all going to be noticeable for simple math.
Part of me wonders whether it is sincerely possible to violate someone's constitutional rights "in good faith". I feel like being a functional police officer should entail knowing the limits on your power, in the same way you're obligated to know what laws you're actually enforcing.
I'm sorry, correcting a story with recent factual information doesn't strike me as undermining the fundamental credibility of a source. Given that they correctly source and attribute the claim that it wasn't tested to a recent article that was accurate at its time of publication.
And given that Tesla did not respond to their request for comment on the data, nor did the national highway safety administraiton, who was supposed to correct them, how?
And it's an interesting time to think about this problem.
Because we're about to see how all the structures that were put in place to enforce those ethics react to being told they must instead do keyword searches for forbidden phrases set by fiat.
If you're the kind of administrator who does boring work of making sure all professors in your biomedical campus submitted IRB paperwork, are you also going to be the person who makes sure their papers don't contain the word "Woman" or "Historically"?