I think that you might be taking "knowledge" too literally. As I read it, the parent's post was using "knowledge" poetically for the various works that AI companies stole.
You wrote your house's security code on a Post-It note. Somebody came in as a houseparty guest and located the note, and discretely copied it onto their hand to get access to the house during your next vacation.
Which verb would describe their knowledge of this security code?
The knowledge wasn't obtained illegally or illicitly, immorally perhaps.
Illegal entry/trespass is a crime regardless of the method used however.
In your analogy copying the code (online content) isn't necessarily illegal, however using that knowledge without compensation, permission or attribution may be a crime (depending on copyright law, judges rulings etc).
Seeding is often considered the actual point of infringement, aka sharing copyrighted material without the right licence.
I should note I suppose that I don't agree with companies using artists or developers hard work for their own financial gain without compensation, whatever form that may take.
The problem as usual is the availability of legal loopholes when you have a team of lawyers and near infinite money.
It really depends the country, downloading a Hollywood movie via torrent is itself illegal in a lot of countries. The only exception I’m aware of is Switzerland
Whatever it is, slapping "steal" on it because what was done is wrong, and stealing is wrong, is the same sloppy logic that the copyright lobby used in the 80s/90s.
I think this is just another part of the growing pains of working with machine intelligence that we have to endure.
Much like we previously had to cope with "hallucinations" as an issue.
If the ultimate goal of AI is to develop general intelligence, the first big objective is: thinking systematically. And the road toward systematic thinking right now is mainly coding, mathematics, and other "verifiable reward" domains.
Claude doesn't have a separate mind for "coding" and "writing". Claude has tokens, and tokens can be assembled in various productive structures, mainly optimized right now for systematic reasoning. Also, a token isn't just a chunk of text. A token is like a little neural-network subroutine that fulfills a function. The conversion of a token into a piece of text only happens on the output side...
When the model finds token sequences that lead toward better verifiable outcomes, it leans hard into those token sequences, and uses them as an essential component of its thought process. "Load bearing" is load-bearing. "Verify, rather than assume" is a mantra that produces good results, so it gets repeated over and over again.
It's super-interesting that this particular moment, where the idea of "Claudish" has become a full-fledged meme, coincides with such astonishing progress in coding and math. My wife says when she uses Claude, that it feels to her exactly like talking to an autistic Engineer.
Not a coincidence, I think :)
My feeling is that the next big era of machine intelligence will require more lateral-thinking and creativity, and hopefully then the models "writing" will be more pleasant to read.
SO a company Github's size can't handle the slop they invested a lot to enable and this is not a failure? I mean if the market moves and you cannot move with it, what do we call it?
