I think what made the AI bubble was that it had some economic characteristics that economists love. There seemed to be a barrier to entry because the initial investments were extremely high, and it seemed that it scaled with money not labor. Environmentally important: money == energy in this context - ThreadSky

patigallardo.bsky.social • 40 days ago

I think what made the AI bubble was that it had some economic characteristics that economists love. There seemed to be a barrier to entry because the initial investments were extremely high, and it seemed that it scaled with money not labor. Environmentally important: money == energy in this context

Comments

patigallardo.bsky.social•40 days ago

That’s what DeepSeek broke. The «natural» barrier to entry turned out to be false. And the scaling with money… this might be the same as my joke about productivity: with programming level of productivity, after the initial investment things get practically free, which means prices are in a free fall

patigallardo.bsky.social•40 days ago

With dramatically less money/energy and the possibility of local models, this becomes accessible to everyone and the monetization of personal information impossible with efficient local models. This doesn’t solve a lot of other issues, but it does break a lot of the power these people had.

patigallardo.bsky.social•40 days ago

Fingers crossed this will become a tool like spellcheck vs «an industry». And humanity will adapt to incorporate it in their work, and the «value» of the output will be nothing, because generating it will be «free». And humans will still appreciate art because it comes from humans. And that matters.

patigallardo.bsky.social•40 days ago

I don’t know about you, but I don’t want my time wasted by word soup with no real thought behind it or pictures that aren’t a person trying to punch me the right way on the stomach to prove a point.

patigallardo.bsky.social•40 days ago

But if a nurse in a village somewhere can be reasonably confident that driving this patient to a hospital 5 hours away is worth it, because the skin-scanner says this mole is cancer with a 98% confidence, then I think that’s a win.

patigallardo.bsky.social•40 days ago

We shouldn’t forget, tho, that the broligarchs were ready to burn down the planet to feed the bullshit generator with energy we really couldn’t afford to spare.

olafurw.com•40 days ago

"The barrier to entry is very high"
Barrier manufacturers: "Excellent"
"Someone invented a ladder"
Barrier manufacturers: "How is this fair?"

azariah.bsky.social•40 days ago

I just heard that in your voice :D

gearon.org•39 days ago

Following your initial post, but going in a different direction than the rest of the thread, another of those economic characteristics that economists and executives love is the commoditization of work. Executives would love to not need to employ people, and AI offers a path to that

gearon.org•39 days ago

Personally, I love AI.
It can scale on tasks that people can't do, or can't do without getting bored/distracted. It can evaluate consistently, which can make it an invaluable assistant in pattern recognition tasks like medical diagnoses. It's great for finding connections in data at scale.

gearon.org•39 days ago

But instead, it's being applied in all sorts of ways that we don't want, and don't need. It's not creative, but it's being used to write essays, books, and screenplays. It's being used to create images and music. And after the initial novelty is gone, people can tell… and they hate it. I hate it.

gearon.org•39 days ago

The idea that AI can be used to replace people's jobs is a pipedream from people who don't understand what today's AI really is, and what its limitations are. Maybe AGI could do it, but we're not there, regardless of what OpenAI says to Microsoft (thereby allowing them to cut Microsoft off)

gearon.org•39 days ago

Investors think that dumping large quantities of money and energy into AI will create something that will appeal to those who would rather not hire people anymore. It's the wrong motivation, and even if the technical developments work out, it's taking us in the wrong direction

patigallardo.bsky.social•39 days ago

Very much, and partially that’s where the weird «productivity» thing comes from: when the cost of making something goes to zero the value of the thing also goes to zero.

lonjil.xyz•40 days ago

Wait, what does that have to do with economists?

patigallardo.bsky.social•40 days ago

Classic economic theory considers the contributors to production to be capital and labor. Capital is needed for initial investment into production equipment (think factory, machinery) and raw materials (think wood, parts, metal).

patigallardo.bsky.social•40 days ago

What happens when there is no factory? No raw materials? Just humans doing things with their brains? The need for capital is gone.

Comments

Posting Rules

Reply