Skip Navigation

InitialsDiceBearhttps://github.com/dicebear/dicebearhttps://creativecommons.org/publicdomain/zero/1.0/„Initials” (https://github.com/dicebear/dicebear) by „DiceBear”, licensed under „CC0 1.0” (https://creativecommons.org/publicdomain/zero/1.0/)E

Encrypt-Keeper

@ EncryptKeeper @lemmy.world

Posts
6
Comments
1435
Joined
3 yr. ago

  • ChatGPT 5 power consumption could be as much as eight times higher than GPT 4 — research institute estimates medium-sized GPT-5 response can consume up to 40 watt-hours of electricity

    Jump

  • AI models require a LOT of VRAM to run. Failing that they need some serious CPU power but it’ll be dog slow.

    A consumer model that is only a small fraction of the capability of the latest ChatGPT model would require at least a $2,000+ graphics card, if not more than one.

    Like I run a local LLM with a etc 5070TI and the best model I can run with that thing is good for like ingesting some text to generate tags and such but not a whole lot else.

  • ChatGPT 5 power consumption could be as much as eight times higher than GPT 4 — research institute estimates medium-sized GPT-5 response can consume up to 40 watt-hours of electricity

    Jump

  • I mean no not at all, but local LLMs are a less energy reckless way to use AI

  • Battlefield 6 players are crying out for a 'real' server browser, and it's about time we demanded the basic FPS feature that Call of Duty killed

    Jump

  • I know exactly why they didn’t make dedicated servers and why doing so would be a scramble. But we are going to need them regardless

  • It’s 2025, Why Are We Still Getting Multiplayer Games Without Crossplay?

    Jump

  • The way it usually works out in the type of games I play (like Battlefield) is that console plays with aim assist get a slight advantage to things like medium range target acquisition and shooting on the move, whereas whereas long range gunplay like sniping favors MKB, and CQC actions like quickly turning up to 180 degrees in short range heavily favors MKB.

  • It’s 2025, Why Are We Still Getting Multiplayer Games Without Crossplay?

    Jump

  • Tell that to my PS5 K/D ratio. Never underestimate the power of aim assist lol

  • Deleted

    Permanently Deleted

    Jump

  • A $10VPS would not be sufficient for a heavily used multi-user Nextcloud instance, and it wouldn’t come with enough storage either.

    You could cloud host this thing for less absolutely, but not a whole lot less. I have a Vultr VPS (cheaper than Digital Ocean, Linode, and other cheapo VPS providers) and all it does is reverse proxy and do some caching and it’s scraping by at a total of $24 a month. A $40 solution that’s more functional if not over-engineered for the difference in price equivalence to a Netflix subscription is not that huge a deal. Especially for someone who is doing this level of work for their job. That $16 is chump change.

  • As electric bills rise, evidence mounts that data centers share blame. States feel pressure to act

    Jump

  • PSE&G told some coworkers of mine their bill would go up by “as much as 20%” shortly before they went up by 150%. One of them got a bill for $800 for their two bedroom apartment

  • Spotify to raise prices in September

    Jump

  • You mean replacing a bunch of people with AI didn’t work?

  • Deleted

    Permanently Deleted

    Jump

  • This is an awesome helpful comment

  • Deleted

    Permanently Deleted

    Jump

  • I’m pretty sure that’s the point.

  • Deleted

    Permanently Deleted

    Jump

  • Folks in IT. This is one of those “deploy something enterprise grade because you can” type of scenarios. It’s like asking why somebody would play a dry milsim game like Arma when Call of Duty exists. This will cost you more than a simple VPS on a platform but it wouldn’t exactly break the bank either.

  • Buy HDD on sale now or wait for Black Friday?

    Jump

  • Black Friday is a decent deal if you’re buying a larger volume of drives. If you’re only planning on buying a few, you don’t have to wait for to. That being said, a ln unimpressive sale is better than no sale.

  • Tucson City Council votes 7-0, unanimously to kill controversial Data Center

    Jump

  • Ah yes, just stop using the internet.

  • The AI bubble is so big it's propping up the US economy (for now)

    Jump

  • Thats like saying ride hailing and food delivery is not profitable because Uber is not profitable in the US.

    Uber is profitable and has been for years now. They also never faced the insurmountable challenges that AI companies do today.

    work in a profitable AI company and can list you a hundred more.

    No you don’t and no you can’t. If you could, you would have done so by now.

    No startup is profitable - thats by design because profit seeking is not what makes your company successful.

    Startups generally have a plan and realistic path to profitability, unlike the AI companies of today who are not profitable, and have no concrete plan to become so. The “profit seeking” investment phase is what startups survive on until they reach profitability. But many (most) startups do not do so and go bust. The same will happen to AI if they don’t become profitable.

    You may continue to live in your fantasy world based on nothing but hope and strong feelings, but you’ve failed to educate anyone here on anything besides your own ignorance. You are free to stick your fingers in your ears and your head in the sand.

  • The AI bubble is so big it's propping up the US economy (for now)

    Jump

  • Of course I used the company that is the market leader in AI as an example that AI companies are not profitable you donut, that’s how that works.

    They’re not the only AI company that’s not profitable, like I said none of them are. You can take your pick if you don’t like OpenAI as an example.

  • The AI bubble is so big it's propping up the US economy (for now)

    Jump

  • Delusion? Ok let’s get it straight from the horse’s mouth then. I’ve asked ChatGPT if OpenAI is profitable, and to explain its financial outlook. What you see below, emphasis and emojis, are generated by ChatGPT:

    —ChatGPT—

    OpenAI is not currently profitable. Despite its rapid growth, the company continues to operate at a substantial loss.

    📊 Financial Snapshot

    • Annual recurring revenue (ARR) was reported at approximately $12 billion as of July 2025, implying around $1 billion per month in revenue.
    • Projected total revenue for 2025 is $12.7 billion, up from roughly $3.7 billion in 2024.
    • However, OpenAI's cash burn has increased, with projected operational losses around $8 billion in 2025 alone

    —end ChatGPT—

    The most favorable projections are that OpenAI will not be cash positive (That means making a single dollar in profit) until it reached 129 billion dollars in revenue. That means that OPENAI has to make more than 10X their annual revenue to finally be profitable. And their current strategy to make more money is to expand their infrastructure to take on more customers and run more powerful systems. The problem is, the models require substantially more power to make moderate gains in accurate and capability. And every new AI datacenter means more land cost, engineers, water, and electricity. Compounding the issue is that the more electricity they use, the more it costs. NJ has paved the way for a number of new huge AI datacenters in the past few years and the cost of electricity in the state has skyrocketed. People are seeing their monthly electric bills raised by 50-150% in the last couple months alone. Thats forcing not only people out of their homes, but eats substantially into revenue growth for data centers. It’s quite literally a race for AI companies to reach profitability before hitting the natural limits to the resources they require to expand. And I haven’t heard a peep about how they expect to do so.

  • The AI bubble is so big it's propping up the US economy (for now)

    Jump

  • The revenue of AI lies in mass surveillance and ads. But even going full dystopia, that has not been enough to make AI companies profitable.

  • The AI bubble is so big it's propping up the US economy (for now)

    Jump

  • AI as a technology is so far not profitable for anybody. The hardware AI runs on is profitable, as might be some start ups that are heavily leveraging AI, but actually operating AI is so far not profitable, and because increasingly smaller improvements in AI use exponentially more power, there’s no real path that is visible to any of us today that suggests anyone’s yet found a path to profitability. Aside from some kind of miracle out of left field that no one today has even conceived, the long term outlook isn’t great.

    If AI as a technology busts, so does the insane profits behind the hardware it runs on. And without that left field technological breakthrough, the only option to pursue to avoid AI going completely bust is to raise prices astronomically, which would bust any companies currently dependent on all the AI currently being provided to them for basically next to nothing.

    The entire industry is operating at a loss, but is being propped up by the currently abstract idea that AI will some day make money. This isn’t the “AI Hater” viewpoint, it’s just the spot AI is currently in. If you think AI is here to stay, you’re placing a bet on a promise that nobody as of today can actually make.

  • The AI bubble is so big it's propping up the US economy (for now)

    Jump

  • I mean we haven’t figured out how to make AI profitable yet, and though it’s a cool technology with real world use cases, nobody has proven yet that the juice is worth the squeeze. There’s an unimaginable amount of money tied up in a technology on the hope that one day they find a way to make it profitable and though AI as a technology “improves”, its journey towards providing more value than it costs to run is not.

    If I roleplayed as somebody who desperately wanted AI to succeed, my first question would be “What is the plan to have AI make money?” And so far nobody, not even the technology’s biggest sycophants have an answer.