> But this example hilariously contradicts the claim that "the system is stacked against the accused"
How does it contradict that? It sounds like you're saying "The system is, in fact, stacked in favour of the accused", but I'm not seeing that here.
In principle, the system _should_ be stacked in favour of the accused - it's a high bar to prove beyond reasonable doubt that a crime occurred, using only legally obtained evidence. But the examples from the article seem to suggest that in most cases, police and prosecutors have it much easier than that.
Hahaha, that's awesome. A nice adjunct to "Why Don't We Just Kill the Kid In the Omelas Hole" [1] (to take another example of a riff on a classic speculative fiction piece, that exposes the concepts to modern sensibilities).
> you must surely be guilty of something. Otherwise, all offers would be meaningless and worth nothing.
You don't have to be "guilty" of anything to be liable in civil law (which contract law is a part of). "Guilt" is a concept from criminal law. It isn't required for contracts to be enforceable.
In general (there are exceptions) offers alone aren't enforceable and don't result in a contract. You need other elements (agreement by the parties, plus something done in return for what's offered) for a contract to be formed - and then it's enforceable.
> > Text generated by an LM is not grounded in communicative intent
> This means exactly that no representation should exist in the activation states about what the model wants to tell, and there must be only a single token probabilistic inference at play.
That's not correct. It's clear from the surrounding paragraphs what Bender et al mean by this phrase. They mean that LLMs lack the capacity to form intentions.
> You are doing the big error that is common to do in this context of extending the stochastic parrot to a non scientifically isolated model that can be made large enough to accomodate any evidence arriving from new generations of models.
No, I'm not. I haven't, in fact, made any claims about the "stochastic parrot". Rather, I've asked whether your characterisation of AI researchers' views is accurate, and suggested some reasons why it may not be.
> For years, despite functional evidence and scientific hints accumulating, certain AI researchers continued to claim LLMs were stochastic parrots: probabilistic machines that would: 1. NOT have any representation about the meaning of the prompt. 2. NOT have any representation about what they were going to say.
But did any AI researchers actually claim there was no representation of meaning? I thought generally, the criticism of LLMs was that while they do abstract from their corpus - ie, you can regard them as having a representation of "meaning" - it's tightly and inextricably tied to the surface level representation, it isn't grounded in models of the external world, and LLMs have poor ability to transfer that knowledge to other surface encodings.
I don't know who the "certain AI researchers" are supposed to be. But the "stochastic parrot" paper by Bender et al [1] says:
> Text generated by an LM is not grounded in communicative intent, any model of the world, or any model of the reader’s state of mind.
That's a very different objection to the one antirez describes - I think he's erecting a straw man. But I'd be happy to be corrected by anyone more familiar with the research.
gtk-vector-screenshot (<https://github.com/nomeata/gtk-vector-screenshot>) will do this, but for GTK apps only. It relies on a custom protocol layered on top of X Window, and I think traverses the tree of GTK widgets to create a vector representation. For a general screenshot program to work, I imagine it would need some sort of hook into every GUI framework used on your system.
How does it contradict that? It sounds like you're saying "The system is, in fact, stacked in favour of the accused", but I'm not seeing that here.
In principle, the system _should_ be stacked in favour of the accused - it's a high bar to prove beyond reasonable doubt that a crime occurred, using only legally obtained evidence. But the examples from the article seem to suggest that in most cases, police and prosecutors have it much easier than that.