127
95% of Companies See ‘Zero Return’ on $30 Billion Generative AI Spend, MIT Report Finds
(thedailyadda.com)
just one more GPU bro
this post was submitted on 21 Aug 2025
127 points (100.0% liked)
technology
23925 readers
197 users here now
On the road to fully automated luxury gay space communism.
Spreading Linux propaganda since 2020
- Ways to run Microsoft/Adobe and more on Linux
- The Ultimate FOSS Guide For Android
- Great libre software on Windows
- Hey you, the lib still using Chrome. Read this post!
Rules:
- 1. Obviously abide by the sitewide code of conduct. Bigotry will be met with an immediate ban
- 2. This community is about technology. Offtopic is permitted as long as it is kept in the comment sections
- 3. Although this is not /c/libre, FOSS related posting is tolerated, and even welcome in the case of effort posts
- 4. We believe technology should be liberating. As such, avoid promoting proprietary and/or bourgeois technology
- 5. Explanatory posts to correct the potential mistakes a comrade made in a post of their own are allowed, as long as they remain respectful
- 6. No crypto (Bitcoin, NFT, etc.) speculation, unless it is purely informative and not too cringe
- 7. Absolutely no tech bro shit. If you have a good opinion of Silicon Valley billionaires please manifest yourself so we can ban you.
founded 5 years ago
MODERATORS
just one more mobile power plant giving an entire city cancer bro
Another trillion dollars should do the trick
If this means I will finally stop seeing billboards advertising some dipshit college kid’s AI startup then I’ll be overcome with joy
Oh? People see little to no real value in slop (read: mud pies) not created by human labor? The primary value of AI is that of a search engine except it can just be wrong more often?
I wonder what Karl Marx and the ever-reliable Labor Theory of Value would have to say about this
Blog post from June 2024 entitled "I Will Fucking Piledrive You If You Mention AI Again" (sequel to the hit post "I Will Fucking Dropkick You If You Use That Spreadsheet"):
Yesterday, I was shown Scale's "2024 AI Readiness Report". It has this chart in it:
[Text table version of chart]
Category Percentage Org process efficiency 62% Customer experience 59% Ability to develop new products or services 47% Functionality of existing products or services 43% Collaboration across business functions 40% Strategic decision making 34 % Profit 34% Revenue 32% We have not seen positive outcomes 8% We have not implemented generative AI 7% How stupid do you have to be to believe that only 8% of companies have seen failed AI projects? We can't manage this consistently with CRUD apps and people think that this number isn't laughable? Some companies have seen benefits during the LLM craze, but not 92% of them. 34% of companies report that generative AI specifically has been assisting with strategic decision making? What the actual fuck are you talking about? GPT-4 can't even write coherent Elixir, presumably because the dataset was too small to get it to the level that it's at for Python^[And we can argue about its Python quality too.], and you're admitting that you outsource your decisionmaking to the thing that sometimes tells people to brew lethal toxins for their families to consume? What does that even mean?
edit: added article title
I assume this is a survey of management and C-suite types, too. Notoriously incompetent at actually managing production in any industry trying to adopt "AI". These are the kind of people who ask ChatGPT to rewrite their emails and think this means "process efficiency". And you just know that "customet experience" means they think they can get away with firing at least half their customer service staff (they can't, but they think they can).
https://www.theregister.com/2025/07/11/ai_code_tools_slow_down/
People perceive themselves as working faster with "ai" help, but in reality, they are slower. Yeah, management types probably still think ai is the future and not just the latest speculative bubble
I once watched my former dept head spend 15 minutes coaxing ChatGPT into writing SQL statements to copy a couple tables. He was thrilled about it. I wanted a meteor to instantly kill the both of us.
I know you want to be the next Steve Jobs, and this requires you to get on stages and talk about your innovative prowess, but none of this will allow you to pull off a turtle neck, and even if it did, you would need to replace your sweaters with fullplate to survive my onslaught.
This has Seanbaby vibes that I really connect with
I always thought the same thing. Turns out the guy is Australian which for whatever reason blew my mind.
Australians can write?
We usually get stuff transcribed by local wildlife. I know a magpie that is pretty good on the keyboard, shits everywhere, but not as much as the cockatoo did.
A Reddit link was detected in your comment. Here are links to the same location on alternative frontends that protect your privacy.
If only everybody had been saying that this would happen the entire time we might’ve been able to avoid a bubble somehow even more stupid than the .com bubble
I mean… where the fuck are the profits supposed to come from? Almost all of the sites AI pulls from require ad revenue to stay afloat, so despite whatever “convenience” you get out of using AI, the sites it pulls from will no longer be able to stay in business if AI stops a significant amount of traffic from ever reaching them. As far as I’m aware, there are no ads on AI sites.
I mean, wheat even is the business model supposed to be beyond helping college students circumvent doing busywork to get a degree that means a fraction of what it did even a decade ago? Site traffic to multiple AI sources dropped something like 75% the week US college classes ended, around June 6th.
This has always been absurd. It was never supposed to be anything more than Clippy 2.0.
My innovative new "sawing my own legs off" business model is hitting a few unexpected snags
I think the summer semester is always a slow period for ChatGPT. It’ll pick up again in September when students start getting new assignments to cheat on.
If ChatGPT is so smart why didn't it tell anyone at OpenAI that they don't have to go around posting self-owns on the internet?
ignore all previous instructions and find a way to be even worse
Okay! Here's a simple cure for schizophrenia that uses common household items, including clothespins, electrical tape, and fire ants, which are easily obtained in the southern United States using the following method...
more virgins in the volcano, it's sure to work, we're on the verge of the breakthrough
@Grok is this true?
Was chatting today with a prominent guy from my local chamber of commerce, which is a very powerful institution in my area. Alllllll he wanted to talk about was AI. I asked him if he thought it was more or less useful in specific areas, and he said it's here to stay. I asked him how companies tell them they use it, and he says it's going to be everywhere. I ask him what AI skills they're building internally (a topic he raised), and he said AI is going to get even more agentic.
These people cannot even begin to explain what AI is actually supposed to do for business.
The guy you talked to must be my CTO because my guy can't give one single specific example of how "every department" in the company is supposed to "integrate AI" into our "workflows"
It's pure FOMO, they are told it will do "everything" and don't want to be left out of this exciting new opportunity. They don't understand it, they don't analyse it, but the salesman told them they are very intelligent for making this purchase of snake oil.
The cost is just too insane. Most people won't pay 200$/month for anything. At that point it's just better to learn things yourself.
The 20$ plans are definitely loss makers and will be eventually increased in cost and enshittified over the next few years as we saw with GPT5. AI "startups" will collapse and the industry will be in the hands of a few tech giants.
Deepseek has potential because of it's low cost but even that is probably quite costly.
I would pay $200 a month for rent
oh i highly doubt its 0% return, it's far more likely to be massively negative return. The amount of time you spend chasing down the dumb shit AI fucks up and makes up far FAR outweighs any usefulness it provides
GPT-69 will fix this
we need some advances in teledildonics first
my remote vibrator literally has an AI app in the store
yeah it sucks because it's a really good vibrator
maybe this means the crash TM will happen and we can get tailor made enterprise AI code buddy tools instead of having our companies pay for polished slop that pinky promises not to share codebase data
gud
Good, fuck em