this post was submitted on 04 Jul 2025
115 points (100.0% liked)

TechTakes

2027 readers
243 users here now

Big brain tech dude got yet another clueless take over at HackerNews etc? Here's the place to vent. Orange site, VC foolishness, all welcome.

This is not debate club. Unless it’s amusing debate.

For actually-good tech, you want our NotAwfulTech community

founded 2 years ago
MODERATORS
top 21 comments
sorted by: hot top controversial new old
[–] kamenlady@lemmy.world 16 points 8 hours ago* (last edited 8 hours ago)

they look forward to turning chatbots into a sea of spam:

We expect rapid adoption of advertising models, transaction fees, affiliate revenue, and marketplace models.

We're doomed.

In the last weeks Pinterest became unusable imo. The AI "sea of spam" is no joke. 7 in 10 posts are ads now. AI ads. Every one of them is a grotesque AI mimic of the content you're viewing, all words meaningless gibberish. The things on the thumbnails suggest, but you can't make things really out by just seeing the thumbnails.

So i clicked them a few times too much. First by curiosity, then by mistake, because Pinterest does everything to make an ad look like a post.

7 in 10 posts.

After all these years successfully procrastinating with Pinterest, it has become a dopamine blocking experience.

[–] salacious_coaster@infosec.pub 27 points 11 hours ago* (last edited 11 hours ago) (3 children)

For reference, the "Hopeless Dipshit Percentage" in any population is about 25-33%.

About a quarter to a third of the population believes in witches, ghosts and ESP; that the earth revolves around the sun; that Saddam Hussein was involved in 9/11; that Obama was born in Kenya; and that evolution and climate change are hoaxes. A third of the US population can't name a single right guaranteed by the constitution or even one branch of government. And a quarter of the population self-professes that they wouldn't stop supporting Trump no matter what he did.

In that context, only 3% willing to pay any money for AI is an utter failure. The LLM bubble needs to burst yesterday, and the whole Internet needs to roll back to 2022.

[–] truxnell@aussie.zone 1 points 3 hours ago

I would far prefer the rollback to be 2004 or earlier...

[–] dragonsidedd@mstdn.science 11 points 11 hours ago (1 children)

@salacious_coaster @dgerard Wait till you learn the percentage of ostensibly fully functional adults that have an Invisible Friend in the Sky

[–] salacious_coaster@infosec.pub 16 points 11 hours ago (1 children)

Yeah but that particular lunacy is grandfathered in to society, so I avoid it in my rants

I don't know how true this still holds but two years after 9/11 70% of the US thought Saddam Hussein was involved in 9/11. My guess is that number continues to be higher than 30%.

[–] NoneOfUrBusiness@fedia.io -1 points 10 hours ago

I mean, "ideally" (to AI companies) those 3% would be the people who use it the most, so businesses and employees who get real value out of the stuff. Depending on who are considered AI users, it's not awful as a B2B thing. Selling to the general public is definitely a no-go though.

[–] Evil_Shrubbery@lemmy.zip 6 points 9 hours ago

That is known & not a problem for the megacorps.

They are building an environment where using AI will be a must (like smartphones that spy on you have become today).

At that point it becomes overpriced & will under-deliver (the monopolistic enshitification of an already shitty offer).

[–] HedyL@awful.systems 19 points 12 hours ago (2 children)

Completely unrelated fact, but isn't the prevalence of cocaine use among U. S. adults considered to be more than 1% as well?

(Referring to this, of course - especially the last part: https://pivot-to-ai.com/2025/06/05/generative-ai-runs-on-gambling-addiction-just-one-more-prompt-bro/)

[–] fullsquare@awful.systems 8 points 8 hours ago (1 children)

cocaine users are convinced that everyone not using cocaine will be left behind, as use of cocaine leads to increase in productivity and innovation

[–] dgerard@awful.systems 6 points 8 hours ago
[–] kamenlady@lemmy.world 6 points 8 hours ago (1 children)
[–] dgerard@awful.systems 1 points 4 hours ago

when I put a pull request in my vein and I tell you things aren't quite the same when I'm rushing on my run and I feel just like Altman's son

[–] zout@fedia.io 11 points 11 hours ago

Seems a lot to be honest.

[–] angstylittlecatboy@reddthat.com 7 points 11 hours ago

epic

Enshittify faster pls

[–] peteyestee@feddit.org 10 points 12 hours ago (1 children)

They are going to make it sub service to kill off the poor people from living with society.

[–] Airowird@lemmy.blahaj.zone 10 points 12 hours ago

While still scouring the free web parts for training data...

[–] brap@lemmy.world 6 points 11 hours ago

I’ve used it here and there for recipe inspiration based on what’s in my cupboard, but really don’t see any other use for it my life. I would drop it in an instant if it became chargeable because it’s pretty shit at most things otherwise.

[–] pineapple_pizza@lemmy.dexlit.xyz -2 points 11 hours ago

I would pay for AI for personal use. But TBH the free models are more than enough for my needs already, so there's no reason to pay for something more advanced. Also, often these "more advanced" models are slower. I'd take speed over some wall of text that takes a while.

We're in the SliceLine era of AI. Enjoy it while it lasts.

[–] cabron_offsets@lemmy.world 0 points 12 hours ago (1 children)

I use ChatGPT to answer the questions for my annual mandatory idiotic work safety training. Just copy/paste the questions and choices in, boom, get the right answers, don’t even have to read the shit. I’d pay $0.01 for that.

[–] kamenlady@lemmy.world 7 points 8 hours ago

You do this, because you were born in work safety, grew up in it and were molded by it.

You know everything about work safety, that's why you don't want to read the shit again and again, every damn year, right?

[Natalie Portmann look of concern]

Right?