“Bad actors” are abusing open source decomps in “increasingly troublesome ways”…

1 comment

sorted by: hot top controversial new old
[–] 8 points 3 weeks ago
22 comments in c/pcgaming@lemmy.ca view that post
[–] 61 points 3 weeks ago (15 children)
  • [–] [S] 42 points 3 weeks ago (3 children)

    There's a lot to hate about AI but for me, it's like the owner of the site says, it's the fact that people don't use AI as a tool but use it to produce crap because the people who are using it has no clue about the software and techniques behind it and have no desire to learn. We're losing institutional knowledge and the passions behind it, which is not good in the long term.

  • source
  • parent
  • hideshow 3 child comments
  • [–] 23 points 3 weeks ago (1 child)

    Losing knowledge, discouraging future experiments from future "junior" developers, and AIs relying on that knowledge fixed in time is the worst that can happen to software development. I'm usually a positive guy, but I don't see how it can be good for the future.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 2 points 3 weeks ago

    Yeah...I really like AI, it's such a fun tool to build things with. It's like magic

    But I don't know how we're going to make new senior devs like this. The baby coders are being deprived of opportunities to go through all the trials by fire that give you the intuitive understanding of how software works

  • source
  • parent
  • [–] 16 points 3 weeks ago (9 children)

    In a statement posted on the site’s homepage, @encounter notes that while they don’t believe AI is the main issue in itself, because when used properly and ethically it “can expedite the decompilation process in a significant way”, the “community aspect of decompilation has been broken” by people using AI in more questionable ways.

  • source
  • parent
  • hideshow 9 child comments
  • [–] -2 points 3 weeks ago (8 children)

    If the only justification for continued use of a tool is that it would be great in a hypothetical world where people don't use it incorrectly, then there's no justification for continued use of the tool.

  • source
  • parent
  • hideshow 8 child comments
  • [–] 1 point 3 weeks ago* (7 children)

    I guess this is a philosophical issue, to me it works the other way round.

    It's not the tech that's the issue, but it is broader democratic social organizational forms that need reform to make adoption of tech (LLM/genAI or other technology) or a wide range of other issues a lot more equitable.

    LLM/ML tech is the future, it's not like the crypto/NFT scams, not even close.

  • source
  • parent
  • hideshow 7 child comments
  • [–] 4 points 3 weeks ago (4 children)

    In any size group of people there will be some who do one thing, and some who do something else. If you're waiting for everyone to use AI correctly in order for its use to be worth it, you'll be waiting forever. The statement is correct when it says that AI technically isn't the main issue in and of itself - the issue is actually with humans as a species, but humanity isn't going to change - there will always be a significant amount of people who use any given tool incorrectly.

  • source
  • parent
  • hideshow 4 child comments
  • [–] 1 point 3 weeks ago (3 children)

    This may be true, but forms of social organisation can and do change and improve.

    It's not possible to put the genie back into the bottle, so to speak, when it comes to technology.

  • source
  • parent
  • hideshow 3 child comments
  • [–] 4 points 3 weeks ago (2 children)

    When the genie costs an immense amount of money to use, and produces things that aren't worth the cost, it doesn't need to be put back- it will be abandoned once we realize we can't make money off of it.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 1 point 2 weeks ago (1 child)

    And what if the full costs are partially (or even mostly) externalized onto someone else?

  • source
  • parent
  • hideshow 1 child comment
  • [–] 1 point 2 weeks ago

    That's exactly the thing - the reason anyone besides the wealthy is able to use AI at all is because they're letting us use it at a heavily discounted price in an attempt to figure out how to make it profitable. And they're failing.

  • source
  • parent
  • [–] 1 point 3 weeks ago (1 child)

    Butnit is. None of these companies are making any money. Are people going to pay whatnis required foe the frontier models to be profitable? And local is no where near on the same level.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 1 point 2 weeks ago* (last edited 2 weeks ago)

    They are not making any money because because the legal/economic/social system of the US doesn't require profitability at this stage, that comes later once they achieve a monopoly-like position and they switch to more explicit rent-seeking behaviour and more direct corruption methods.

    FWIW, we probably agree on much more than this short convo would imply. I am just looking at things from a different angle.

  • source
  • parent
  • [–] 8 points 3 weeks ago*

    Ironically, this is even happening to LLM inference engines.

    Llama.cpp and other community projects are drowning in manpower shortages.

    Yet dozens of vibecoded “forks” keep popping up, to support some specific thing and fragment the community. There are also a number of commercial “Tech Bro leeches” that sap attention from the main project and send support tickets their way, while contributing precisely nothing back, with Ollama being the most notorious.

  • source
  • [–] 5 points 3 weeks ago (4 children)

    My thoughts there are the AI tools brings about a whole new challenge that they are just completely walking away from, and that’s fine, but to say community is lost is absurd. It sounds like their community exploded and they decided it was too hard to collaborate now that there is 10x the number of participants.

    New challenges that need to be solved are automation frameworks for ensuring quality of the decomp, automated testing to ensure faithful reproduction of bugs, etc. tons of more things that need to be done.

    So walking away when they pioneered the community is one thing, if they want to leave the next chapter for someone else to pick up the reigns, but refusing to acknowledge the new community is quite frankly disheartening.

  • source
  • hideshow 4 child comments
  • [–] 6 points 3 weeks ago (2 children)

    The bit about vibecoded forks sapping attention is really important, though.

    It’s different than the usual “solo dev fork drama.” Vibeforks spawn in the dozens, often go out of the way to draw attention, fizzle out quickly and leave precisely nothing constructive behind. They actively drain the main project.

    It seems like a small thing, but these kinds of projects gain manpower and motivation through exposure. I’m seeing this exact same thing erode communities in other niches, too.

  • source
  • parent
  • hideshow 2 child comments
  • [–] -3 points 3 weeks ago (1 child)

    My point is that this is the reason to popularize a foundational framework that these vibe coders can pull from and have a baseline quality to them. Standardized license, skeleton project plan, etc. Then even if they are abandoned by the original, it could still be useful if further development is needed and it should meet a minimum quality standard.

  • source
  • parent
  • hideshow 1 child comment