Show newer

@ct_bergstrom

I'm sure all those exploited workers who were payed $1.25 to think about this for five minutes were carefully considering the philosophical implications of their answers. I guess you just have to put "ChatGPT" in the title and you get a pass through peer review.

BTW, Google offers a utilitarian answer too. I was hoping for some effective altruism but it let me down.

@ct_bergstrom So good that a third ski is just waiting to protect your fractured tibia!

@ct_bergstrom Yeah, that joke didn't really work, did it? "I aim at the stars (but sometimes I hit London)"

@ct_bergstrom Both ChatGPT and GPT-4 compensate for Google by underestimating the Rosy Wolf Snail by 1 or 2 orders of magnitude. But the geometric mean is pretty close to the right answer!

This is proof that they're talking and AGI is here. Cross posting on Twitter now.

Joe boosted

If you're into astronomy, this is just so wild - a black hole ejected from a galaxy created a 200,000 light-year long trail (larger than our own galaxy) of new stars. hubblesite.org/contents/news-r

@freakonometrics The real headline - I asked ChatGPT to code a handlful of specific code snippets - isn't quite as clickbaity.

Joe boosted

@Popehat Note that the student petition didn't even bother to mention two of the most common PTSD categories: motor vehicle accidents and traumatic injury/illness/death of a loved one.

Joe boosted
Joe boosted

The fact that ChatGPT is still just straight up making up information about *people* and even including fake citations just blows my mind. The defamation suits are already starting (reuters.com/technology/austral) and it's not like there's a Section 230 for LLMs. I guess this was optimistic of me but I totally expected a "band-aid" fix by now, like "I'm sorry, but as a large language model I am incapable of providing references." Two more recent examples I saw in thread:

@ct_bergstrom Just trust the companies! That's the kind of ethics advice you get from one of leading figures in the field.

The guy's becoming the Elon of AI research.

venturebeat.com/ai/10-years-on

@jetjocko Just substitute "nuclear weapons" for AI in that screenshot. Remarkable how arrogant some of these folks are.

Joe boosted

What if, instead of destroying humanity or becoming our god-emperor, all a self aware AI wanted to do was open up a nice cafe and catering outfit?

Show older
Qoto Mastodon

QOTO: Question Others to Teach Ourselves
An inclusive, Academic Freedom, instance
All cultures welcome.
Hate speech and harassment strictly forbidden.