2011
The erasure of Luigi Mangione (substack.evancarroll.com)
submitted 2 years ago* (last edited 2 years ago) by to c/technology@lemmy.world
260 comments fedilink hide all child comments
 

Right now, on Stack Overflow, Luigi Magione’s account has been renamed. Despite having fruitfully contributed to the network he is stripped of his name and his account is now known as “user4616250”.

This appears to violate the creative commons license under which Stack Overflow content is posted.

When the author asked about this:

As of yet, Stack Exchange has not replied to the above post, but they did promptly and within hours gave me a year-long ban for merely raising the question. Of course, they did draft a letter which credited the action to other events that occurred weeks before where I merely upvoted contributions from Luigi and bountied a few of his questions.

you are viewing a single comment's thread
view the rest of the comments
[+] -6 points 2 years ago (16 children)

Yeah, AI has become good enough at this point that you can provide it with a large blob of context material - such as API documentation, source code, etc. - and then have it come up with its own questions and answers about it to create a corpus of "synthetic data" to train on. And you can fine-tune the synthetic data to fit the format and style that you want, such as telling it not to be snarky or passive-aggressive or whatever.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 16 child comments
    •
  • [–] 24 points 2 years ago (15 children)

    It can reproduce an api. Can't solve actual problems. LLMs are completely incapable of innovation.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 15 child comments
    •
  • [–] 7 points 2 years ago (4 children)

    That's not really true though.. They come up with brand new sentences all the time.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 4 child comments
    •
  • [–] 1 point 2 years ago (3 children)

    No, they can only take from things in their models.

    Moreover, all of them use statistics, typically Bayesian, to get the results. What you get from an LLM is essentially an average* of the model data. This is why feeding LLM output into a model is so toxic, it's already the average.

    • Yes I know it's not really the average, but for laymen us good enough comparison.
  • permalink
  • fedilink
  • source
  • parent
  • hideshow 3 child comments
    •
  • [–] 2 points 2 years ago (2 children)

    They only take from the statistical distributions of words in the context of preceding words (which is why they never say "the the" etc, why the grammar is nearly always correct). But that doesn't mean that whole sentences are lifted from the source material. There are near infinite paths through those word distributions, and many have never been produced by humans, so LLMs do produce sentences that have never been uttered before.

    They couldn't produce new conceptual context spaces in the way that humans can sometimes, but they can produce new combinations within existing context spaces.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] -5 points 2 years ago (8 children)

    And yet the synthetic training data works, and models trained on it continue scoring higher on the benchmarks than ones trained on raw Internet data. Claim what you want about it, the results speak louder.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 8 child comments
    •
  • [–] 1 point 2 years ago (3 children)

    The results aren't worth the expense. So-called "AI" is the biggest bubble since the great recession.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 3 child comments
    •
  • [–] 1 point 2 years ago (2 children)

    Nah, I'm still giving that one to the blockchain. LLMs are going to be useful for a while, but Ethereum still hasn't figured out a real use, and they're the only ones that haven't given up and moved fully into coin gambling.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 1 point 2 years ago (3 children)

    This is the peak, though. They require new data to get better but most of the available new data is adulterated with AI slop. Once they start eating themselves it's over.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 3 child comments
    •
  • [–] 2 points 2 years ago (2 children)

    You are speaking of "model collapse", I take it? That doesn't happen in the real world with properly generated and curated synthetic data. Model collapse has only been demonstrated in highly artificial circumstances where many generations of model were "bred" exclusively on the outputs of previous generations, without the sort of curation and blend of additional new data that real-world models are trained with.

    There is no sign that we are at "the peak" of AI development yet.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •