fosstodon.org is one of the many independent Mastodon servers you can use to participate in the fediverse.
Fosstodon is an invite only Mastodon instance that is open to those who are interested in technology; particularly free & open source software. If you wish to join, contact us for an invite.

Administered by:

Server stats:

11K
active users

No one, and I mean NO ONE wanted to listen to us when we screamed at the top of our lungs that the emperor has no clothes...

Nothing else has taught me more about how propaganda works and spreads across these institutions, whether it is the Nobel Prize, the United Nations, all these governments and whatnot, and how the stock market is basically like gambling, as seeing how everyone treated OpenAI, Deepmind, and friends since their inception.

mempko

@timnitGebru It's so true. Thanks for speaking truth to power.

It's amazing that people who don't know math simply have billions to waste. Seems a bit irresponsible for a society to allow that, especially since there are so many important problems out there.

@mempko Here the academics themselves are in on the grift and misleading the public 😡

@timnitGebru Do you think it's a grift or are they self delusional? Or maybe both?

@timnitGebru I talked to a very smart friend of mine who is in this space. I told him how current approaches (LLMs and friends) are a dead end and we need to go down a different path. Told me I'm too late and AGI will be here in a couple years. That's why I don't think it's all a grift for researchers, I think some drank their own kool-aid.

Or maybe to be the best grifter, you have to believe your own magical thinking.

@mempko @timnitGebru
Maybe scientists don't really care about accuracy as long as the research is fun and drives money. String Theory worked the same.

@Teratogenese @mempko Honestly can't be interested in the stuff they're doing even w.r.t. it being fun and interesting. I was interested in designing circuits because it was like a puzzle. Algorithms were the same. Whatever this is that they're doing is like "what happens if we throw unlimited data and compute?"

@timnitGebru @Teratogenese @mempko or those who said "let's do this N-month research on evaluating/comparing LLMs" to end up stating the obvious 👉 "there's no good metric or a solid way of evaluating the system or what was called hallucination "!

Tons of papers stating the obvious!

@mempko@fosstodon.org I'm watching this play out as researchers I'm in the halo of are actively banking on LLMs to eliminate the cost of hiring data labellers. Their reasoning is motivated by the fantasy of cost savings, not better results.

Watching them bend over backwards to excuse the LLM's middling success rate and to handwave away how it misses almost everything of actual interest to their research is embarrassing.

@mempko @timnitGebru
People may start out just being in on the grift, but they soon start to delude themselves. Most people don't like to think they are souless money grubbers. So they justify and rationalize