AI is both overhyped crap and a revolution.
this post was submitted on 15 Apr 2025
1270 points (96.1% liked)
memes
14201 readers
4054 users here now
Community rules
1. Be civil
No trolling, bigotry or other insulting / annoying behaviour
2. No politics
This is non-politics community. For political memes please go to !politicalmemes@lemmy.world
3. No recent reposts
Check for reposts when posting a meme, you can only repost after 1 month
4. No bots
No bots without the express approval of the mods or the admins
5. No Spam/Ads
No advertisements or spam. This is an instance rule and the only way to live.
A collection of some classic Lemmy memes for your enjoyment
Sister communities
- !tenforward@lemmy.world : Star Trek memes, chat and shitposts
- !lemmyshitpost@lemmy.world : Lemmy Shitposts, anything and everything goes.
- !linuxmemes@lemmy.world : Linux themed memes
- !comicstrips@lemmy.world : for those who love comic stories.
founded 2 years ago
MODERATORS
AI, in some form, is here to stay, but the bubble of tech companies shoving it into everything will pop at some point. As for what that would look like, it would probably be like the dot-com bubble.
Yeah.. tech companies need to chill out with "jumping on the bandwagon" and actually innovate
AI is here to stay imo. This is not crypto
Did you see crypto going away anywhere? Because I sure as hell didn't.
I did. Crypto was never user friendly for the average person to use and the way the US gov requires you to report each transaction on your taxes makes it more trouble than it is worth.
AI on the other hand is extremely user friendly and genuinely is of use to the people I know who use it.
load more comments
(1 replies)
load more comments
(1 replies)
Don't forget "THE CLOUD" and "IoT"
The cloud was the dumbest hype because it changed nothing. Network services were now named cloud services, that was it.
We also got:
- cyber: the realisation that the internet exists but 20 years too late
- big data: companies snorkeling your data for targeted advertising and government surveillance
- blockchain: safe ways to pay your dealer and investment scams
- VR: the weirdest sort of corporate meetings and some cool game experiments
IOT seems kinda ok in that regard. Connect your microcontrollers to the internet to allow some extra coordination.
load more comments
(1 replies)
I do feel that, unlike Crypto, AI (or, to drop the buzzwords, LLMs and other machine-learning based language processors and parsers) will end up having a place in the world.
As it is NOW, the AI hype train is definitely an investment bubble and it will definitely explode in a glorious fashion eventually. Taking a lot of people down with it.
But unlike Crypto, AI does -- It like does things, you know? Even if I personally feel like it's mostly only good for a toy, all my attempts to use it for anything society would deem "valuable" were frustrated, but at least I can RP with it when my friends aren't available. It is a thing that exists and can be used.
Crypto was funny because it was literally useless. Just an incredibly wasteful techno-fetishistic speculative vehicle with precisely zero shame about being that.
As for what's next, I think Quantum Computing might be it. That is, assuming the Tech Industry even survives the bubble's burst in its current form. Because everyone in the industry is putting all their eggs including theoretical eggs that haven't even been laid, and in fact there's not even a chicken in this AI hype train. And even with AI becoming part of people's lives, as I predict it indeed will, when the bubble does burst it might end up hitting the reset button on who is truly in charge of things.
You are talking about crypto after the hype. But actually the idea before the hype was pretty cool. It was basically to allow safe payments around the world without any government control needed.
In my city there were a few shops that took Bitcoin back then and a cup of coffee was something like 2 Bitcoin.
You make it sound as if they don't already have a place in the world. Ml models have been employed to solve problems for the greater part of a decade or more now. Deeply integrated into damn near everything that you interact with.
When you get an MRI or a CAT scan AI helps identify and call out peculiarities.
The traffic lights and traffic management in your city is probably partially operated using "AI".
Wear and tear on parts of your car are predicted from data using ml models.
Industry sensor data is interpreted and made actionable using ml models.
Telecommunication Network fault prediction and detection.
Energy load prediction.
....etc
But you're probably talking about is recent hype around llms which are models that are fantastically good at understanding language. Which opens up a whole new field of possibilities when you can combine the ability to understand language with the predictability and reliability of "classic" ML models.
load more comments
(1 replies)
This is exactly the same at the dot.com boom and bust. After the crash the internet didn't go anywhere, and look at where we are now. The same will (unfortunately) happen with AI.
load more comments
(1 replies)
You're assuming there will be a next time. When the AI bubble bursts, and it will, the whole economy will go down with it. AI companies are massively in debt and have a product that ranges from utter shit to kinda okay, and absolutely no sane way to monetize it. Everyone outside of tech, you know, the customers, fucking hate AI. It has stolen their work, jeopardized their livelihoods, wasted their resources and made the most insufferable asshats in history very wealthy.
The bubble is not going to burst, but it is going to deflate as expectations are adjusted to reality. AI is revolutionary, but it's not AGI, and it will take decades to be fully realized, just like the internet.
Don't most modern tech companies actively lose money and rely on investors to actually pay their workers? Then again AI uses a crapton of resources compared to decade optimized search querying.
load more comments
(1 replies)
load more comments
(1 replies)
I hate that we call any algorithm that gets information by looking at data "AI." If people consider something like linear regression (a supervised model) to be "AI", then "AI" isn't going to pass. Hell, even neural networks are just a shit ton of addition and multiplications.
All computing is just shit tons of math operations.
That's the "artificial" part of "artificial intelligence", so I'm not really sure what you expect AI to look like.
I'm not a big fan of LLMs and I don't think they're intelligent, but if you're disqualifying them based on using math then nothing is ever going to satisfy you
All computing is just shit tons of math operations.
I agree, and "computing" is a great umbrella term for all math operations. And there's a reason you used the term LLM instead of AI, and that's because LLM better describes what you're referring to. The name reflects the function or the most defining characteristic of what you're referring to.
The way people throw around 'Artificial Intelligence' feels wrong to me. The words "Artificial Intelligence" suggest these models are conscious or sentient, which they’re not, so the term ends up being misleading.
So while it’s not technically wrong to use "AI" as a catch-all for anything data-driven, I don’t think it’s nearly as useful or accurate as more specific terms like LLM.
Also, when I hear people use the term AI, it’s usually by those who have no idea what they’re actually talking about. It’s always in the vague, buzzword-y context of “we need to AI our processes” even though, realistically, most systems already have some form of "AI" baked in.
Edit: It's just a huge buzzword that's starting to lose meaning to me: https://www.youtube.com/watch?v=-qbylbEek-M&t=26s
load more comments
(3 replies)
NFTs were just star registries. Pay a fee, and you can claim to own a certain star.
load more comments
(4 replies)
Oh, it's gonna be so much worse. NFTs mostly just ruined sad crypto bros who were dumb enough to buy a picture of an ape. Companies are investing heavily in generative AI projects without establishing a proper use case or even its basic efficacy. ChatGPTs newest iterations are getting worse; no one has a solution to hallucinations; the energy costs are astronomical; the entire process relies on plagiarism and copyright infringement, and even if you get by all of that, consumers hate it. AI ads are met derision or revulsion, and AI customer service is universally despised.
This isn't like NFTs. It's more like Facebook and VR. Sure, VR has its uses, but investing heavily in unnecessary and unwanted VR tools cost Facebook billions. The difference is that when this bubble bursts, instead of just hitting Facebook, this is going to hit every single tech company.
load more comments
(41 replies)
Quantum computing, probably.
Problem is, it has the potential to be actual reality. Tech bros need their products to be 99% blue-sky hype to get their financing, and they can't risk some nerd going "well actually what you're suggesting can't be done any more efficiently on a quantum computer than you can do now".
load more comments
(6 replies)
Reminds me of Blockchain
According to new research from Deloitte, 74 percent of large companies (with sales over $500 million) see a “compelling business case” for blockchain technology.
Indeed, from supply chain management and regulatory monitoring to recruiting and healthcare, organizations are applying blockchain to their business models to revolutionize how they track and verify transactions.
It's not a fake or fundamentally useless technology, but everyone who doesn't understand it is rushing to figure out how they're gonna claim to use it.
Yeah, when someone just describes blockchain, saying "I guess we could use it for supply chain tracking or healthcare tracking or whatever" is a reasonable first impression.
The problems show up the second you start thinking about how to actually implement the damn thing. You don't need a blockchain for logistics or healthcare tracking. It has no inherent advantage over regular databases. It doesn't solve organisational issues. It's just a slow trustless distributed append-only database. It's good when you need a trustless distributed append-only database! Most people don't need one.
Same thing with AI technologies, just a bit different in that it's somewhat more useful. They're good and useful technologies and they have plenty of perfectly valid usecases. Then the tech bros started going "Maybe we could use AI for some weird wacky obscure niche and charge a lot of money for it?" or "we're going this wacky crap whether you want it or not, we don't care what it's necessary for us to do to make it happen, and we'll charge a lot of money for it".
You don't need a blockchain for logistics or healthcare tracking.
https://hbr.org/2022/01/how-walmart-canada-uses-blockchain-to-solve-supply-chain-challenges
load more comments
(5 replies)
Another banger from lemmites
Mate, you can use AI for porn
If literally -nothing- else can convince you, just the fact that it's an automated goon machine should tell you that we are not going to live this one down as easily as shit like NFTs
load more comments
(17 replies)