this post was submitted on 30 Oct 2026
599 points (99.3% liked)

memes

23217 readers
3007 users here now

Community rules

1. Be civilNo trolling, bigotry or other insulting / annoying behaviour

2. No politicsThis is non-politics community. For political memes please go to !politicalmemes@lemmy.world

3. No recent repostsCheck for reposts when posting a meme, you can only repost after 1 month

4. No botsNo bots without the express approval of the mods or the admins

5. No Spam/Ads/AI SlopNo advertisements or spam. This is an instance rule and the only way to live. We also consider AI slop to be spam in this community and is subject to removal.

A collection of some classic Lemmy memes for your enjoyment

Sister communities

founded 3 years ago
MODERATORS
 
you are viewing a single comment's thread
view the rest of the comments
[–] bagsy@lemmy.world 21 points 1 day ago (2 children)

Sorry, but sellers will NEVER let the price come down. It will take years, if ever that prices start coming down again. Also, there is almost zero chance the rich give up on their dream of replacing expensive employees with GPUs. Even if the market corrects a bit and washes out some weak businesses, the AI build out isnt going any where. There is money to be made replacing employees, and someone is going to make it, come hell or high water.

[–] CheeseNoodle@lemmy.world 5 points 1 day ago (1 children)

As someone in the midst of being replaced by AI, it ain't happening. It's superficially great at my job but the moment anything needs fixing or changing its more work than if I (or a human in general) had been paid to make it properly in the first place. No real productivity has been gained or time saved, those costs have all just been kicked down the road. The only thing LLMs (specifically LLMs, narrow/specialist systems have legitimate uses) have really enabled is the ability to produce Bethesda levels of technical debt in record time.

[–] motruck@lemmy.zip -4 points 21 hours ago* (last edited 21 hours ago) (3 children)

Sorry bro. It is over. AI may not solve all the problems they say it does but it can code and they'll continue to get it better at coding. Software is now something you can imagine and create without writing it.

[–] elephantium@lemmy.world 8 points 21 hours ago

Wishing a thing to be true does not make it so.

[–] mushroommunk@lemmy.today 6 points 21 hours ago

I don't know why people keep trying to tell us senior software engineers LLMs can code. I've rejected enough garbage pull requests to know it isn't so.

[–] Equinox1289@sh.itjust.works 2 points 20 hours ago

AI can only be as good as its available training. The only improvements have been clever ways to approach that asymptote, but no true work arounds. To get them to truly learn, would require algorithms better than diffusion and backpropagation, and those currently do not exist.

[–] Jiral@lemmy.world 18 points 1 day ago* (last edited 1 day ago) (1 children)

"Even if the market corrects a bit" I think you have no concept on how large that bubble has become already. This could heavily impact the real economy.

After the bubble implodes it will become much harder to finance such data centres, never mind that it will be much harder to find a place to build it, even with all the corruption in the world.

[–] bagsy@lemmy.world 4 points 1 day ago (2 children)

I understand the numbers are truly staggering, but I've also been around the block enough times to know employers and investors will go to unbelievable lengths to save costs on labor. If the 1% believes AI is going to cut labor costs, then nothing will stop them. Nothing.

[–] TrackenDaKracken@lemmy.world 4 points 1 day ago (1 children)

Elon said it. The big companies with other sources of revenue, like SpaceX, Google, Microsoft, and Facebook will survive. It's the myriad of startups that are neck deep in debt for building out data centers that will fall.

And, while eventually AI might fulfill its promise, it's not a given. It may not be possible with any amount of money and effort, like faster than light travel.

[–] Equinox1289@sh.itjust.works 2 points 20 hours ago

With the current tech, AGI is literally impossible. They believed scaling, clever architecture, and context window tricks would solve the issues of the current LLMs, but fundamentally diffusion and backpropagation cannot iteratively learn as they have no way to train during inference.

[–] Jiral@lemmy.world 4 points 1 day ago

That doesn't matter if the house walls are coming down because the cards are falling. Of course there is something that will stop them, the real cost of LLMs. There are no real costs yet, they are hidden by the bubble. The 1% loves cutting labour cost but not when that means losing money.