• 0 Posts
  • 1.15K Comments
Joined 3 years ago
cake
Cake day: October 6th, 2023

help-circle



  • LLMs absolutely have a market. It’s a market that’s probably in the low tens of billions of dollars worldwide. Non-critical translation, content writing, image generation, mockup. Etc etc. there are absolutely uses for LLMs and other generative AI.

    The problem is that hundreds upon hundreds of billions have been invested, with hundreds more on the way. The current investment will never be recovered.

    And these aren’t durable investments, the equipment has a very limited shelve life, the chips aren’t going to last nearly long enough to recover their cost, and the data centers are already obselete for the next generation.

    Edit: getting downvotes for recognising that there is in fact a market for generative AI, by people who didn’t catch that “the low tens of billions” is about 1% of what AI boosters are proclaiming.














  • Tar_Alcaran@sh.itjust.workstoScience Memes@mander.xyzSpicy Air ☢️
    link
    fedilink
    English
    arrow-up
    3
    arrow-down
    1
    ·
    edit-2
    20 days ago

    So let me counter ask you a very similar question: how much radioactive material (weight or volume, your choice) do you think was spread in Chernobyl,

    Some 60 tons of reactor fuel were expelled “locally”. That wasn’t easy to Google, but easily to convert back from the radiation released. I might be a bit high due to iodine being released which isn’t part of the fuel.

    Thanks for once again proving my point. As soon as I point out how nuclear waste isn’t actually a real problem, opponents of nuclear power tend to immediately move the goalposts, without actually answering the question too.

    But the preemptively adress your moved goalpost:

    That might be flippant, but does this matter at all? You might as well say solar panels are deadly because some idiot didn’t tie his safety line while installing rooftop solar panels. Or some DIYer wired the electrics wrong and burned their house down. People have died from solar panels, so using your logic, solar panels might at any moment strike and kill someone!

    It doesn’t work like that. Solar panels are entirely safe when used properly. Nuclear is entirely safe when you don’t intentionally build a gigantic bombs and then intentionally push it past all limits and override all safeties. No electricity reactor before or after Chernobyl has been capable of failing this way, it was literally uniquely terrible.

    Since you also didn’t answer, for everyone who actually does care: since 1957 till today, humanity has created, from all sources of nuclear power generation, about 260.000 tons of spent nuclear fuel. If you were to stack it into a single pile, it would form a 23m cube, or cover a soccer field 2m deep. For ALL fuel ever. And we could reprocess all of it, if not legal opposition to it.

    That’s the amount nuclear opponents complain about. One 23m cube in 70 years of power generation.



  • Tar_Alcaran@sh.itjust.workstoScience Memes@mander.xyzSpicy Air ☢️
    link
    fedilink
    English
    arrow-up
    4
    arrow-down
    1
    ·
    edit-2
    20 days ago

    Solar is cheaper, even at high latitudes like in northern europe, even for baseload application with big battery buffers right next to the solar farm.

    Honestly, that sounds extremely unlikely. I don’t live that far north in europe, and while I manage about 0kWh on my residental panels on a yearly basis. Thanks to seasonal changes, I would either need 4 more rooftops to keep the power on during january, or I would need to bank something like 700kWh to make it through 3 winter months. That’s not counting the electric car, or heating. Heating would roughly quadruple the numbers (being almost entirely clustered when solar isn’t producing), and the car would add roughly another house on top (assuming 50% is charged away from home).

    Quick maths that I did because I wanted to try going off-grid: I would need ~100m2 of solar panels, and 2500kWh of battery storage. Or on a national level, 63 TWh of storage as well as just under a 1000km2 of solar panels if everyone lived as low-footprint as we do. And that’s just housing, it doesn’t include commercial buildings or industry.

    The big buffers next to the solar farm are actually quite tiny. The largest under-construction battery park in the netherlands banks about 1200 MWh. With an average househould consumption, that’s just about enough to carry some ~4000 2-person households through the 3 winter months, assuming you put down enough solar to meet your yearly energy household energy demands (which we don’t have). They’re obviously not meant for long-term storage, but long-term storage is exactly what you need to make solar work.

    And nuclear doesn’t have any of these issues. The only issue is that it’s expensive, because we stopped building them.


  • Tar_Alcaran@sh.itjust.workstoScience Memes@mander.xyzSpicy Air ☢️
    link
    fedilink
    English
    arrow-up
    3
    arrow-down
    1
    ·
    20 days ago

    Because of the long build-time, you can buy the batteries 10 years from now, comparing to a nuclear plant that starts construction today.

    Sure, but that’s a shitty comparison, because I can also build 20 nuclear powerplants, and bring costs WAY down. And that’s the thing. These comparisons are always “If we keep boosting X, and supressing Y, then X will perform better!”. Yes. Duh.

    Look at what China is doing. They’ve built dozens of plants in the past years, and have >30 under construction right now, with ~150 planned. They’re building them for a fraction of the cost, because they’re not completely reinventing them every single time.

    It’s not like I am saying we should scrap ongoing constructions.

    Fair, we shouldn’t. But my worry is that even in 10 years, we’re still going to be using lots of fossil fuels, and that will always be lower if we ALSO build nuclear. Or at the very least stop heavily opposing it.