OpenAI now has 400M weekly users. I think the conversations that I still see that AI overall is not useful or just a bubble are not reflecting reality. There is a lot more importance now in discussions of how to steer the use AI to improve things (& mitigate risks) than arguing that it doesn’t work - ThreadSky

emollick.bsky.social • 8 days ago

OpenAI now has 400M weekly users. I think the conversations that I still see that AI overall is not useful or just a bubble are not reflecting reality.

There is a lot more importance now in discussions of how to steer the use AI to improve things (& mitigate risks) than arguing that it doesn’t work

Comments

bakerblue.bsky.social•8 days ago

Agree, ardent dismissals amount to denial. The uses it has been put to so far that focus on creating things like text to-video seem weird-and maybe threatening to a lot of folks..the average person is likely to see this tech as threatening-until it is shown to benefit them.

mokoringo.bsky.social•8 days ago

"Is not useful" is obviously wrong, but couldn't it be that the bubble/hype is around the fact that many people think it will replace humans? Like whoever says SWEs will not be needed when it is the exact opposite - they will be even more needed, just working differently and using AI too

wavebidder.bsky.social•8 days ago

yeah, unlike cryptocurrency, which is an end run around banking regulation, there's real value, like with the https://dot.com bubble

spiralizing.bsky.social•8 days ago

Whats the statistics on the "type of use"? Like how many of those accounts use it for not replies, scams etc vs productivity?

emollick.bsky.social•8 days ago

There are lots of surveys of use. It varies even within person, cheating at school mixed with actual learning; roleplaying and helping with advice; etc

spiralizing.bsky.social•8 days ago

Is there any statistics on how that number compares with scams and bot replies?

phillipcarter.dev•8 days ago

And also, the importance of how peoples' perceptions of AI are largely shaped by ChatGPT and its default settings.

nicksallon.bsky.social•8 days ago

Extraordinary to think that the *but it hallucinates* argument is still so prevalent given how powerful capabilities are

skewballfox.bsky.social•8 days ago

Honestly, at least for AI agents, the fact that it hallucinates at all is a major limiting factor.

sgnt.ai•7 days ago

Strong disagree, and I work full-time with LLMs as a job. LLM hallucination is a major nuisance, both in terms of just chatting to ChatGPT on one hand, and then in terms of powering systems based on LLMs.

sgnt.ai•7 days ago

It's a huge problem, and we have at best partial solutions. That doesn't mean it can't be used for real-world things, but you can't just hand-wave away hallucination as if it's an occasional thing, rather than a systemic flaw that needs to be managed

nicksallon.bsky.social•7 days ago

Yes - its a problem and needs to be solved, but using it to dismiss AI and it’s capabilities is a mistake

loganmischler1436.bsky.social•8 days ago

Seems like there are three competing narratives: AI is going to kill us all. AI is going to usher in a utopia. AI isn't going to do much of anything, other than maybe help us code and write emails faster.

ddelalamo.bsky.social•8 days ago

Are you getting a sense from folks who pay the full $200 price tag for Deep Research & Operator that it is worth the cost?

sgnt.ai•8 days ago

And yet there was that terrible Ed Zitron article recently trying to claim that people only use it because it's in the media a lot

tq.tqwhite.com•8 days ago

Ed Zitron is a hack. I've watched him interviewed several times. His viewpoints are never based on valid observations. He's a living outrage engine.

Comments

Posting Rules

Reply