He missed a bit:

they will, rightfully so, switch to a different instance … or go somewhere else entirely

then this isn’t a problem for the broader fediverse community to solve

This is the natural end result of every volunteer run instance, you don’t find it odd that over the last 40 years of the internet not one fediverse like server or community has survived or even been mildly popular?

I’ll repost this because for some reason the other post got deleted, it was regarding lemm.ee shutting down, they were concerned that one of the largest Lemmy instances is shutting down and the future of Lemmy:

You’re 100% right to be concerned and to be honest I have doubts lemmy will ever crack more than a few million users, the same thing happened with Mastodon, something that relies so heavily on volunteers running the infra almost inevitably results in burnout because the fediverse works on a disincentive basis:

Basically the more popular a server is, the more funding it requires, the more admins it requires, the more work it requires, and all of this is on a slim margins or more likely requiring on people to donate time/money/effort ‘for free’ is a huge ask.

The supply of people sitting around doing nothing all day who care enough to dedicate their time/effort/money to running a social network… for free… is a very small group, almost as small as the amount of people who are willing to donate every month to a social network.

You can find mods of communities are usually fans of the communities they mod, it’s a topic they enjoy and so the incentive for them to invest their time is to keep their community clean and great. But running a social network which has hard costs not just time is a whole other thing

This is opposed to a regular website or social media network, where as it gets bigger, it makes more money through ads/subscriptions, the incentive is to get bigger to make more money

And then they can simply pay people to do the shit no one wants to.

The reality for me is that the money has to come from somewhere, you can do a paywall like newspapers do or beg for donations every page visit like the guardian/wikipedia do, or the usual suspect allow advertising, but the money has to come from somewhere.

Thus the fediverse has a disincentive to growing larger, it is simply easier and more sustainable to remain small

I don’t have sources but I think it makes sense logically.

pretty sure most women in a polygamy hate being in them, read more than a few articles about first wives then getting put aside for a younger model but they can’t leave because no money (based islam 🙄)

I have no idea how you use this and stay sane but 🫡 to you sir

very easy to get started, make sure you have a graphics card with ideally more than 6gb of vram (the more the better), grab lm studio: https://lmstudio.ai/ then under the discover section you can grab a local model

these ones run locally on your PC and don’t touch the internet hence the more VRAM you have the faster they go and more larger models you can run

the opposite! ai is great for helping learn a language, part of my goal is to get away from english and us centric everything, if i just continue to watch other cultures in english i’m not doing anything really except watch more english language content

Sam Altman will face the same fate as Elizabeth Holmes.

You might want to pick someone else besides her, her product was fraud, openai is used by millions of people every day

You’re 100% right to be concerned and to be honest I have doubts lemmy will ever crack more than a few million users, the same thing happened with Mastodon, something that relies so heavily on volunteers running the infra almost inevitably results in burnout because the fediverse works on a disincentive basis:

Basically the more popular a server is, the more funding it requires, the more admins it requires, the more work it requires, and all of this is on a slim margins or more likely requiring on people to donate time/money/effort ‘for free’ is a huge ask.

The supply of people sitting around doing nothing all day who care enough to dedicate their time/effort/money to running a social network… for free… is a very small group, almost as small as the amount of people who are willing to donate every month to a social network.

You can find mods of communities are usually fans of the communities they mod, it’s a topic they enjoy and so the incentive for them to invest their time is to keep their community clean and great. But running a social network which has hard costs not just time is a whole other thing

This is opposed to a regular website or social media network, where as it gets bigger, it makes more money through ads/subscriptions, the incentive is to get bigger to make more money

And then they can simply pay people to do the shit no one wants to.

The reality for me is that the money has to come from somewhere and subscription wise you can do a paywall like newspapers do, or beg for donations every page visit like the guardian/wikipedia do, but they money has to come from somewhere.

Thus the fediverse has a disincentive to growing larger, it is simply easier and more sustainable to remain small

DeepSeek is an improvement, but not to the level initially reported.

🫠 I cannot be any clearer:

previously ai model training was entirely based on simply buying more chips as fast and as hard as possible, deepseek changed that

Perhaps because you didn’t understand what they said.

they repeat what i said, did you read them? previously ai model training was entirely based on simply buying more chips as fast and as hard as possible, deepseek changed that

From your own article

Is it impressive that DeepSeek-V3 cost half as much as Sonnet or 4o to train? I guess so. But OpenAI and Anthropic are not incentivized to save five million dollars on a training run, they’re incentivized to squeeze every bit of model quality they can. DeepSeek are obviously incentivized to save money because they don’t have anywhere near as much.

https://www.seangoedecke.com/is-deepseek-fast/

The revelations regarding its cost structure, GPU utilization, and innovative capabilities position DeepSeek as a formidable player.

https://www.yahoo.com/news/research-exposes-deepseek-ai-training-165025904.html

^ fyi that article you linked to is an AI summary of a semianalysis.com article, maybe AI is useful after all ;)

If the models were actually getting substantially more efficient, we wouldn’t be talking about bringing new nuclear reactors online just to run it.

Youtube uses a fuck ton of power but is an incredibly efficient video delivery service

The growth and popularity of AI and its uses is simply outpacing the efficiency gains

why are you linking me to articles i read ages ago?

Nor is smartphone AI going to do the things people what AI to do. It won’t let the CEO take your job.

You think AI is only useful if it’s taking someones job?

The current models depend on massive investment into server farms

I hate to tell you this but your knowledge of AI appears to be limited to 2023 ;)

You missed the entire Deepseek fiasco which basically put an end to the “just order more chips” strategy of AI

Chinese company DeepSeek made waves earlier this year when it revealed it had built models comparable to OpenAI’s flagship products for a tiny fraction of the training cost. Likewise, researchers from Seattle’s Allen Institute for AI (Ai2) and Stanford University claim to have trained a model for as little as US$50.

https://theconversation.com/microsoft-cuts-data-centre-plans-and-hikes-prices-in-push-to-make-users-carry-ai-costs-250932

Or if you’d like to read an absolutely mega article on it: https://stratechery.com/2025/deepseek-faq/

And no, self-hosted models aren’t going to make up for it. They aren’t as powerful, and more importantly, they will never be able to drive mass market adaptation

Both Samsung and Apple have on device AI already, you’ve not seen the Apple ad? https://www.youtube.com/watch?v=iL88A5F9V3k

They’re only planning more and more features using it

They aren’t as powerful

We’ve had insane gains in locallama’s since 2022, including but not limited to this from the other day https://lemmy.world/post/30442991

And every few weeks a newer and improved model comes out, I’ve never seen tech so amazing and progress so fast as I have AI

Like, if you don’t get the problem here, I really don’t think I can explain it.

tbh I might need AI to explain it to me because to me you just said: people like trash

hence the success of a lot of the Marvel films despite rapid decreases in quality, or the Star Wars sequels.

The part I don’t get is, what difference does it make (to you) if the trash is made by a guy using AI or made by a company?

The main thing is that quality movie cinema goers like yourself have an organic free range no ai involved movie experience, and so long as there is enough of you out there (and there should be) then you’ll be able to enjoy your cinematography experiences while the rest of us watch slop (and ultimately isn’t this what is currently happening anyway?)

Not only that but are you assuming that AI and computational power will never reach a point where we’ll be able to create a wide variety of films and movies that are on par with or exceed today’s movies?

Are we assuming there will never be a movie that’s made using AI that is better than these:

https://en.wikipedia.org/wiki/List_of_films_considered_the_worst

or the garbage that’s on TV at 3 in the morning?

a person using AI is telling me they don’t value their own mind to the point they rather outsource it to a dumb prediction machine

Why do you believe this is so?

Most creative work is taking what we think of (an idea or thought) and putting it into action, a movie is just a more elaborate version of a story which we have been telling each other since the beginning of time

If I tell Veo3 a story and it brings the story to life with visuals and audio where have I not valued my own mind?

This feels like elitism: I had to learn a whole craft (singing/song writing/instrument etc) to bring my idea to life, you didn’t use a fancy camera with fancy video editing software, you just told a computer what you’d like to see or hear, how dare you!

The reason why the original hot take is bad is because it is simply referencing all the most obvious limitations of LLM’s at the time (problems with fingers, over explaining answers)

This would be like in 1995 saying smartphones?? like tiny computers that fit in your pocket? Wow that’ll be so shit! Oh great so now my wrist watch computer takes 5 minutes to boot up so I can check the time? and then it’ll probably blue screen of death! I bet I’ll get a virus then have to run a virus scan on it ughhhhhh

Do half of you realise you would get better quality discussion if you weren’t so cunty

But most people will never see one, nor do they care to see one

then who cares?

this is like being sad people don’t care for Marionette shows anymore

so you don’t know how capitalism works

mcdonalds is cheap and largely made by automation, in fact if it was made by automation today most people probably wouldn’t realise … but restaurants and cafes still exist

if there is demand for organic movies with no ai then we’re going to build it

Companies will try to replace as many jobs as possible with AI, meaning everything gets considerably worse in quality and reliability, and there won’t be any new jobs to replace the old ones

lol that’s not how capitalism works

i mean it literally is, this hot take from 2023 is already out of date with veo3

and this is after 2 years, can’t even imagine how incredible it will be in 20 or 30 years

over robust, verifiable information

Random websites are robust, verifiable information now are they? How times change

30 years ago I was told they are unreliable and to use books in the library for research

20 years ago I was told using WebMD was unreliable, after all it will just say you have cancer laugh out loud! Using the internet for medical information? Crazy!

I wonder where we’ll be in 20 or 30 years time