11,937 karma · joined June 5, 2012
It would also help if you could substantiate your initial claim (i.e. "internet training data is not where frontier capabilities come from")
That is only relevant in the US, and even there it is still not clear-cut whether the fair use doctrine applies on all these scenarios. Outside of the US the situation is also quite different: for example take a look at the recent ruling on GEMA vs OpenAI in Germany.
The reality is that the copyright issue with generative AI is very complex and reaching anything resembling a conclusion will take much more than a few opinion paragraphs from an American district judge.
https://www.bloomberg.com/news/articles/2026-06-22/spacex-ki...
Why is that almost every LLM generated article sounds like a LinkedIn motivational post?
(this is not a rhetorical question, I would really like to know why, from all the writing styles, this is the most prevalent one)
https://www.heraldscotland.com/news/26188090.john-swinney-ta...
And in relation to your first comment, most sane people would agree that "tools" don't exist in isolation - neither come into existence out of nowhere.
This reductionist position of treating extremely complex machines with deep social interactions as a tool like any other is objectively wrong, and I believe the reasons are highly obvious but I can expand on this if you disagree.
If GP has access to this dataset it would be interesting to know how sparse is the data in that area.
A tiny bit hyperbolic for someone who's not a fan maybe? :)
You could say the same about MP3 encoders but I don't think that would convince any judge
I also don't remember ever seeing anything being force-fed as much as LLMs, why?
Not that things were amazing before that... it's just that mid 10s had simply too many bad actors showing up at the same time with a lot of money and power.
If you want more examples look into IoT products like smart toothbrushes, many of them now are "AI enabled".
People are just tired of being mislead and abused by corporations, which is why there is now a market for non-tech products.
I was actually just looking into that: Beale Infrastructure, owned by Blue Owl.
we saw a gap: some of the largest global hyperscalers—companies like Microsoft, Amazon, Google—needed partners who could move at their speed and scale to solve their growing data needs
https://wealth.blueowl.com/learnengage/ra/capturing-growth-t...
The CEO is ex-Azure/Microsoft:
https://www.forbes.com/sites/bernardmarr/2025/07/08/microsof...
Apollo caps investor withdrawals from flagship private credit fund
https://www.ft.com/content/d315cb1c-1e1e-479c-a6d4-b3a817fea...
Ares limits withdrawals from $10.7bn private credit fund
https://www.ft.com/content/9a1b60c5-7015-4314-9608-d02c00c3a...
Store birth date in systemd for age verification
Nasdaq's Shame
https://news.ycombinator.com/item?id=27676266
Here is a more recent example I found in Cursor's browser experiment from January: