“watching how a human toddler learns about the world”
I have several family members with kids now and this is quite funny. A toddler learns how to crawl by looking at a lot of adults crawling or something.
“watching how a human toddler learns about the world”
I have several family members with kids now and this is quite funny. A toddler learns how to crawl by looking at a lot of adults crawling or something.
I think he is a cs guy, who also is into EA, so basically it is the same source as the research OP posted about.
It seems like it is a neat addition to a robust verification system, sadly they picked it as a replacement for a verification system. Ah the libertarian desire to build a thing but not be responsible for it.
I dont even get where they are going with this, It is a bit like asking a suspected troll if they are a troll, if they answer yes they are a troll, if they answer no you still suspect they are a troll.
(This is assuming they are not doing critihype, lets ask them. Oh no).
We just are very good at anthropomorphizing. We created pet rocks for example (also showing that capitalism is more than happy to jump into this)
The AGI learned DECIEVE, but all i wanted it to learn is HUG.
Oddly parodied before it happened in the tv series community (the dad of the somewhat racist main character. The dad itself is very racist).
E: I do wonder what Javier Milei looks like if he would dye his hair
The object vs meta level.
Wait, as AI learns the same way as humans, and humans are thus basically AI does that mean I can just pirate everything? This will change… nothing really.
Isn’t this a case of ‘the good bits are not original and the original bits are not good’. According to wikipedia it is from 2005.
I used this site (forgive me for the very 2000’s style branding, very edgy etc) to learn python. the course used to be free on the site, so you will have to find a way around that, either via wallet, or 1337 skills (the course doesn’t do the same branding as the site btw). But it also has a useful list of links to books and stuff like that to learn more (or at least give you an idea about how much different things exist out there).
But the idea behind the course ‘the best way to learn is to do the work’ is pretty useful in learning how to code. It is easy to fall into a trap of reading about some coding and thinking you understand it and then utterly fail at actually implementing it.
But as froztbyte says, it does depend a bit on how you learn.
E: also this url is quite old now, so I have no idea how many of the links still work, sorry about that.
The starcraft apm thing always amused me, people who instead of giving an order once, just keep clicking that mouse and issuing the same move order over and over again because apms. Good way to teach Goodhart’s law to Gamer Brains.
because he refuses to play competitively or follow any of the rules around organized speedrunning, musk’s doing the modern, depressing equivalent of claiming to be the strongest guy around (no you can’t see him lift any weights in a competition setting, only the suspiciously light ones in his home gym)
See also how he claimed Zuck was avoiding him and didn’t want to fight him because he would lose. (yeah, going to Zucks home when he is not home and offered to fight you in a real ring which you keep ignoring makes you the winner really).
Or see his twitter stats. Before the muskening of twitter, twitter kept various public (because publicly traded) stats which people could see, monthly increase in something like monthly active users which can be targeted by advertising, stuff like that. (the growth rate of which was apparently about 1-2% per month, which is quite impressive imho), but now he talks about ‘unregretted user minutes (up by 10% this year(*)), and stuff like that’. He never mentions that (according to the stats I looked into shortly before the takeover) twitter always grew in users, he makes it looks like he did something special. Like a guy buying a restaurant transformed it into a mcdonalds and then goes ‘look we sold a lot more hamburgers than last year’.
*: I mention this because I assume that people can do a bit of math in their head and can compare 1-2% monthly growth with 10% yearly, even if it isn’t the same stats.
He wants to be seen as the Uber-nerd, better at nerding than everybody else, so of course he would cheat. See also how he has claimed he was the best at quake. He just is hype and bravado because a group of people who saw him stutter (*) about some half remembered/understood science fiction ideas were impressed with his genius and drive up his stocks/reputation. He now is going after the anti-woke nerds as potential marks (He has said quite a few dumb thinks about video games recently).
See also how his elden ring build was bad, his diablo 4 world record relied on abusing an exploit, he thinks polytopia is some sort of complex high level game on the level of chess. The man is a dullard. (E: He also is bad at dnd., a cooperative game which you basically cannot fail to play well))
*: Nothing wrong with having a stutter, that happens. It is weird people claim his stutter is not because he just stutters, but because it is a sign his brain is so great that he is having a hard time because it is thinking about so many genius level things at the same time.
The problem with stuff like this is not knowing when you dont know. People who had not read the books SSC Scott was reviewing didnt know he had missed the points (or not read the book at all) till people pointed it out in the comments. But the reviews stay up.
Anyway this stuff always feels like a huge motte bailey, where we go from ‘it has some uses’ to ‘it has some uses if you are a domain expert who checks the output diligently’ back to ‘some general use’.
‘i am a stochastic parrot and so are u’
reminds me of
“In his desperation to have produced reality through computation, he denigrates actual reality by equating it to computation”
(from this review/analysis of the devs series). A pattern annoying common among the LLM AI fans.
E: Wow, I did not like the reactionary great man theory spin this article took there. Don’t think replacing the Altmans with Yarvins would be a big solution there. (At least that is how the NRx people would read this article). Quite a lot of the ‘we need more well read renaissance men’ people turned into hardcore trump supporters (and racists, and sexists and…). (Note this edit is after I already got 45 upvotes).
Btw slur against people who refuse to use ai apparently dropped. “Refuseniks”. A very Bruce noo moment.
Also a very, google that new word you just invented because this is a very weird reference moment: https://en.m.wikipedia.org/wiki/Refusenik (even if the term itself is very grokAI (aka wordplay that 60 year olds think is edgy or cool, but actually so without edge it will not cut warm butter)).
E: whoops wrong thread, imagine this was in the random sneers thread
I was thinking stuff like servant, mom, slave, wife something like that. But would have expected agent instead of thing, as that seems to be the term people go for and put hopes on, even Schneider is on the agent train (but we should make it ethical!)
No problem, set your useragent to openai, boom no more advertisements. Every other video is some weird video about how great google is however. Strange that. (Note: this is a joke, not something that I tested and actually works).