I miss consistent weather.
this post was submitted on 30 Jan 2026
318 points (99.1% liked)
Technology
79674 readers
3986 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related news or articles.
- Be excellent to each other!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
- Check for duplicates before posting, duplicates may be removed
- Accounts 7 days and younger will have their posts automatically removed.
Approved Bots
founded 2 years ago
MODERATORS
The snow keeps melting
A company with a $3.2T market share. The game is made up and the points don't matter.
"market reaction suggests that more capital isn’t going to be a viable substitute for a business model anymore."
Time to find the next vague thing that investors can pour trillions into without really knowing what it is or does.
It'll be quantum computing. Since the last hype around it, a lot of progress has been made to the point that quantum computers are actually becoming useful, since error correction is now mostly resolved.
Robotics. It will be a pivot to robotics.
Which will create a labor boom, to operate the robots, a la Tesla
Don't forget, though, it does that one thing for that one reason I forgot already as I typed it... But its still good, clearly!
Please fucking crash I want to be able to buy basic computing hardware again
Since OpenAI just announced the possibility of bankruptcy, it's definitely coming. It's going to be wild for whichever idiot in charge at MS to go down in history as the man who ruined one of the most powerful and integral companies on earth.
Wait, where? I wanna read and savour it.
idk, it was late last year that Sam said he expected OpenAI revenue to grow steeply, but also that if it doesn't then the company could go bankrupt by 2027 at the latest.
I’m here hoping the bankruptcy of OpenAi happens sooner.
lol they havnt even put the ads in yet
They don't have a product with any actual value or use cases. The ads aren't going to reverse that. If it were that simple then they would have been able to make profit with their subscription model.
How would that even work as well? The ads will be in the website, doesn't most stuff run through API calls? If you force everyone to start paying per call, the business model falls apart instantly.
API calls are already only paid, no?
I'm guessing the ads will be embedded in the answers of the free users (like: it will add to the prompt something like "and don't forget to plug the sponsor, ridge wallet")
I think we’re quite a long way off before they actually crash and burn, if they ever do. We have no idea how much money the ads will inject and they also receive significant government contracts and will probably get a lot more going forward
If the market can pretend Tesla is worth so much i think it can easily sustain AI for many years
It's already been several years. Tesla had an actual product that people wanted. Yes, they've been doing their best of late to torpedo their market share and brand name but at one point they were doing what they set out to do. Open AI has never done what they said they would do.
Kinda but also not entirely. I know a lot of people who use ChatGPT and other AIs at work and it does basically exactly what they want and just gets better
I’m not a proponent but the naysayer doomers are almost as wrong as the tech evangelists
Is it overvalued? Sure
Is it worthless? Absolutely not
That's cool. I have yet to find a use case for AI. Am I doing it wrong or are they just bad with computers?
I am with gustofwind here. I use AI to help me quickly draft up mindless policies, then read through it and edit it where needed. It is a lot faster than any typing I can do.
I also use it to help me find configs for stuff I deploy, but I make sure it attaches the source link for me, so I can read the original source docs and keep a sane approach to what I am doing.
Again, I also think it is way overvalued, but to say it has not helped me build successful stuff over the past 2 years would be a lie. My 2c
Yes, why do you care about the pointless corporate crap you produce . Ai slop is the desired result you are just letting your pride get in the way.
Products and services have been dropping in quality well before ai slop
You overestimate the quality of acceptable work in many industries. AI and a little human editing and oversight is perfectly capable of producing legitimate work product.
The real problem is capitalism driving everything to shit and that really has nothing to do with ai influenced workflows
I went from Windows laptop and Netflix and Hulu to a Linux desktop for a home server running Immich, Mealie, Jellyfin, and the Arr suite in docker containers. All proxied on Cloudflare for remote access. I would never have been able to do that without the use of ChatGPT. I had no knowledge of software development, Linux, networking, etc at all. If you know how to query, AI can be a huge aid in learning. It's helping me brush up on my Italian right now too since I haven't spoken it in 5 years.
Tbh it’s not much different than search engines. You need to learn how to use them and when it’s appropriate to do so…it’s basically a skill issue 🤷♀️
Reminds me of when search engines first arrived and we were taught very early in school how library research works and then when to use digital academic databases vs regular search engines or just hit the books.
And yeah tech support is a great use case and you can just use the Gemini links that send you to the Reddit threads where the information came from to verify it.
I feel like if you’re minimally responsible it’s pretty hard to have AI backfire on you
That's cool. I did all of that without AI coming from a similar place as you. AI didn't open up a new path for you, it just showed you a path that already existed, which isn't any different from what a regular search engine can do. There was nothing stopping you from finding that path on your own except your unwillingness to look.
Willingness to look is a pretty important factor. LLMs reduce the personal cost incurred to look up information, similar to how search engines saved us from having to go to the library for every question we had.
They presented to you a reasonable use case (assisted learning) and your response was "lol, you're just lazy. Do it on your own. I did it, so can you".
I am in a similar position, networking is Martian to me and if I search guides on how to do stuff, it's full of people that go "just use X to do a reverse proxy", as if I have 200h of experience under my belt. I'd rather have a chatbot explain to me like I am 5 in some cases.
Doubt that'll happen for a few mor years unfortunately. I can't imagine most of the hardware made for AI datacenters is compatible with consumer stuff :/
A lot of it hasn't actually been made, though. The AI companies have put in orders for future production. That future capacity can be redirected with a wave of a pen.
OpenAI has made about $1.4 trillion in commitments to procure both the energy and compute it needs to fuel its operations. But its revenue barely crossed $20 billion in 2025.
Investors are increasingly critical of what they describe as “circular” deals involving the industry’s biggest players. On Wednesday evening, The Information reported that OpenAI is seeking a fresh $60 billion in funding from heavyweights like Nvidia and Amazon. However, market reaction suggests that more capital isn’t going to be a viable substitute for a business model anymore. “Maybe Oracle stock got way ahead of fundamentals, and now the market’s saying, ‘All right, show me, I want to see it,’” Eric Diton, president of the Wealth Alliance, told**Yahoo Finance.
Thanks for sharing, really insightful.
In my personal opinion after being also responsible in AI for our company, I do not see how it will be profitable for them.
For example Microsoft Copilot license, costs 30$/month, but a lot of things I can do with it a free Chatbot can do too.
It definitely has it strengths and use cases and I am sure it will not go away. But it is not the way the market it as a full AI, it just generates answers with the highest probability. I cannot see it developing from there to the real AI.
I think this year will be really interesting to watch all the AI companies, especially Oracle as they have to refinance a lot. If one falls it will send them into to a spiral, the big companies will be fine, but I am sure they will cut their funding of OpenAI.
But who knows could be the other way around and OpenAI finds anything new to make them more profitable.
The real lesson here is that if you are a company that was founded on stupid imaginary bullshit your investors are comfortable with investing in stupid imaginary bullshit and it isn't going to hurt your price.
When you are a legacy tech company whose investors expect you to actually make products that you sell for money, they don't like to hear that blew every penny you had on fucking magic beans.
This would be like Big Oil investing in Enron, no?
and so it does
I hate AI...it never was AI. It was useless in all my tests.
The gig was up for me when I tried to get it to play dungeon master in a game of DnD. It would start out great, but eventually it would forget what we were doing and instead of giving me choices it started just telling me the story of me playing dnd and it would stop giving me options. This would happen about 6 minutes into playing, or 3 or 4 "turns", and that's when I realized the incredible memory sync it is if it can't reference instructions given moments ago. A newer model won't fix that.
At the end of the day it's complex predictive text that amounts to a Rorschach test.
You need to do a custom program if you want to do that. I mean a traditional program where variables are stored properly.
The models have no memory at all, at every question it starts from scratch, so the clients are just "pretending" it has a memory by simply including all previous questions and answers in your last query. You reply "ok", but the model is getting thousands of words with all the history.
Because each question becomes exponentially expensive, at some point it starts to prune old stuff. It either truncates the content (for example the completely useless meta ai chatbot that WhatsApp forced down the throat loses context after 2-3 questions) or it uses the model itself to have a condensed resume of past interactions, but this is how it hallucinates.
Otherwise it will cost like $1 per question and more
Which kind of illustrates the fundamental flaw right? Videogame companies have spent decades creating replayable DnD esc experiences that are far more memory efficient and cost effective. They already kind of do it the best way. AI can assist, and things like the machine learning behind the behaviors of the NPCs in Arc Raiders for example is very cool, but as you said, you need a custom program... which is what a video game is, so I guess my point is I don't see the appeal in re-inventing it through sort of automated reverse engineering.
LLMs could theoretically give a game a lot more flexibility, by responding dynamically to player actions and creating custom dialogue, etc. but, as you say, it would work best as a module within an existing framework.
I bet some of the big game dev companies are already experimenting with this, and in a few years (maybe a decade considering how long it takes to develop a AAA title these days) we will see RPGs with NPCs you can actually chat with, which remain in-character, and respond to what you do. Of course that would probably mean API calls to the publisher’s server where the custom models are run, with all of the downsides that entails.
Oh no....
Anyway.
Wait.
You mean that dedicating the majority of their business to buying things from themselves has not proven to be a sound strategy in the eyes of investors?
So it begins
The AI ouroboros if finally consuming itself.
Sucks to suck
Looks like that magic well of social permission is about to dry up.
https://economictimes.indiatimes.com/news/international/us/us-stock-market-crashes-today-why-are-dow-sp-500-nasdaq-down-today-tesla-meta-microsoft-in-red/articleshow/127780471.cms This isn't investment advice, I just thought this might be interesting to read! (Additional angle of same story)