IP is such a stupid concept. Both humans and AI use material to improve themselves and that material is often heavily gatekept (Try getting access to a "regular" scientific paper) and should be freely accessible to anyone. No matter who.
Somebody does something which takes some work. There's nothing special about his work other than that it is available-- there's so many stories you can tell, there's so many ways to express an idea, that to, if the author hadn't created a particular work, making it illegal to create that particular work, people wouldn't have lost anything. It would be a trivial restriction.
Thus when someone takes a particular work and does something with it, he chooses to not develop one of the other similar works he himself could have created, and takes advantage of somebody else's effort.
Because of the variance in production cost and ROI for effort, vs the general ability to copy IP (just the IP, which could be loads of research that isn’t valuable without loads of capital and decades of investment into certain infrastructure e.g. TSMC), there have to be some sort of incentives to do the hard groundwork that creates valuable IP that once discovered is trivial to copy or implement, otherwise people simply won’t do it, they’ll just copy and resell and then fewer people will do the initial leg work until we’re left chasing the bottom of the barrel of the level of effort someone may invest in something they have no means of capturing return from that work and risk. And not everything can be hidden behind a web service, some IP is far more challenging to protect.
So, while I tend to agree that the current implementation of IP protections is trash, I think we do need something there to encourage such work. In research we have(/had?) federal tax investment into doing such work and as a trade off, researchers get fed and live a semi comfortable or sometime decent quality of life for their efforts. I don’t have the answers to the ideal model, I know what we have isn’t great and can be improved upon in some ways, but I don’t know wha the best(TM) solution is here. I know it’s not pure unrestricted markets doing whatever they want, though.
It's not a stupid concept. It exists because someone put effort in. Should we steal their effort?
Imagine if Wikipedia had a paywall for each article… but you had to pay to a different organisation for each vaguely related set of pages. Some requiring recurring subscriptions and minimum spend so you can’t cancel.
Worse, the people demanding the payment aren’t even the operators of Wikipedia and they charge the article editors for the privilege of publishing on their platform!
You can have government funded research at a government funded university and then both the researcher and the students at that same university have to pay to have the research published and accessed!
But not too many pictures please and they can’t have color, that’s expensive, you see?
Raw data hosting? No way! Do that yourself.
Versioning, backlinks, ongoing feedback, comments, etc… don’t even ask! Do you think this is the future!?
Meanwhile the largest collaboration of human minds on the planet takes place on GitHub and is almost entirely volunteers working for free and giving away their work for free for everyone to fork and build on. And comment, contribute, and extend.
Science as it is today is a tyre fire and it is almost entirely the fault of rent seeking middlemen with misaligned incentives.
For IT people, the rough equivalent is why a 1 KB file costs $50/year despite containing not much more than a random number and a text field verified with a single API call.
Sure, you can have a HTTPS certificate for free! You can!
Just not through Azure Key Vault CA automation, because Let's Encrypt couldn't afford the kick-back to a middle manager in Redmond. Or more charitably, because "there's no margin on free".
Never underestimate the ability and willingness of billions of managers to insert themselves into the middle and collect rent on everybody below them.
I'm replying to a comment that negatively compares science publishing to Wikipedia. What stops a scientist publishing work voluntarily on a free platform?