Whenever I say this, I am told the goal of copyright is to incentivize innovation, not protect creators.
So it doesn't matter if every particular person is truly rewarded for their work or if the rewarded person is the one doing the actual work (employers own copyright even though employees do the work). What matters is the impression and the aggregate effect.
And of course if humans not necessary for innovation, it loses its meaning. Copyright is already pretty much dead since many people and organizations get away with running copyrighted work through ANNs and claiming it's not derivative work.
---
But the bigger issue is copyright is only about creativity, not about the human time and effort put in. It doesn't protect most normal work.
Ultimately, every person has a limited time to be alive and that's one area where we're all roughly equal. Even taking skill differences into account, the difference in productivity between people is not that high.
Take a passenger jet and all the work, skill and knowledge that goes into building one. There's no way a single person could do all of that, even ignoring study time as if he magically started with the required knowledge at birth. Yet there are people rich enough to own one. That makes no sense.
> reward people for the full transitive value of their work
> the goal of copyright is to incentivize innovation, not protect creators.
Both statements can be true. IP law attempts to incentivize innovation by instituting a system that is hoped to (very approximately) award creators in proportion to the transitive value of their work. This is much easier to see with patents, where each use of the idea itself requires an explicit license. In the case of copyright it is assumed that both buyer and seller are able to accurately assess the value, that the market is efficient, etc. Generally a bunch of stuff that's only approximately true in the most hand-wavey sense.
> There's no way a single person could do all of that ... Yet there are people rich enough to own one. That makes no sense.
Presumably you own a smartphone. Presumably it was relatively cheap in comparison to your total yearly income. Yet I am quite certain that you could not manufacture an equivalent device from scratch on your own. A modern CPU is hundreds of man years of work just for the blueprint; you still need to figure out fabrication. And then there's the rest of the SoC, the RAM, the display, the radio, ...
Sure. I mentioned it because some people have been pretty hostile to the idea that IP should protect people. They argued as if a technology that makes human innovation obsolete should automatically invalidate copyright because it's would be no longer needed either. And screw the people whose work that innovation was built upon - licenses, consent, etc.
> Presumably you own a smartphone
Maybe the plane was a bad example, I used it because it was one of the first things which made me realize how many orders of magnitude individual wealth spans.
There are everyday items which you could make on your own (e.g. furniture) and on the other end there are massive projects which require their own specialized supply chains (e.g. planes). Smartphones fall somewhere in the middle, probably. They benefit immensely from economies of scale and that the same infrastructure (fabs) can make parts for smartphones, computers and make other device types - both unlike planes.
A more telling comparison perhaps would be how many people you need to get together to make 1) one of the item, 2) how many to make enough to serve that group, and 3) how many so each person in the group has one. A plane can, after all, serve many people at once. Having one for yourself is, in part, where the extravagance of owning one as a singular person comes from.
I may well have been one of those people in a previous exchange. I argue that IP law only protects people as a means to an end. As a severe restriction of individual freedoms, I firmly believe that its only legitimate purpose can be the net benefit of society as a whole. If any given aspect of IP law (including copyright) is no longer required to encourage innovation then it should be abolished.
All of that is perfectly consistent with the ideal of rewarding people for the full transitive value of their work which itself follows naturally from the goal of incentivizing innovation.
Regarding the plane, I apologize if my tone tended towards quibbling. I understood and agree with the point you were trying to make about income inequality. Private jets constitute an almost absurd level of physical resource allocation to the individual. Jet engines alone require significant quantities of rhenium, an element that's slightly rarer than platinum. I don't really see how any of that relates to copyright or AI though.
I think the discussion should move away from freedoms/rights to consent.
Compare with rights in the physical world:
- If I own a property, whether it's a house or plot of land, I can consent or not consent to other people being there. Walking across a field of grass doesn't damage or depreciate is in any measurable way but nobody is allowed to do that unless I allow it.
- If I own a bike I never use and never intend to use, nobody is allowed to take it for theirs, not even borrow it, without my consent.
It should be the same with my work which is not physical. If I publish code online, my consent should be required for using it in any way - especially if i give specific instructions such as a license file.
Without that, those who are best able to capitalize on an opportunity (typically those who are already in positions of power and have enough wealth) gain the most from it, not those who contributed to the opportunity through their labor.
---
And it shouldn't be about innovation either. If i record my cat doing something funny, I can a) keep the video to myself, or b) upload it to youtube where it has a tiny but real chance to hit millions of views for meaningless internet points, ad money and presumably preferential treatment of my future videos from the algorithm, or c) I can upload it to my website where almost nobody will see it. But the choice c) should not allow Youtube to take my video and host it on their platform. Because I don't consent to that. In fact, I should be able to say for example "you can only rehost this video if you include a link to my website and don't make money from it".
And I should definitely be allowed to say "you can't use this video to make a statistical model of all videos on the internet which can then reproduce non-verbatim patterns from them". And by "you can't", I mean "I don't consent to it and society/law agrees with punishing you if you do".
There's nothing innovative about funny cat videos (for all we know, the exact situation has been recorded thousands of times and posted on youtube already and only one of those videos reaches more than a few views by random chance). Innovation shouldn't be relevant, consent should.
---
> All of that is perfectly consistent with the ideal of rewarding people for the full transitive value of their work
Not sure if I understand but I think I agree in a theoretical sense but disagree in a practical one. Theoretically, IP law isn't a requirement for reaching this ideal, maybe there is a different rule system which can achieve that. Practically, IP law is a flawed mess but we shouldn't throw it away until we find a better alternative. (I've seen a discussion here where a person basically argued for destroying copyright but only several posts later revealed he was doing it because he had an alternative in mind. Ironically his alternative didn't require copyright to be abolished, it was perfectly compatible AFAICT but also was opt-in so it was effectively powerless and couldn't work towards his stated goals.)
> I apologize if my tone tended towards quibbling
It's fine, I used the example because it was one of the things which caused cracks in my acceptance/ignorance of how the world works and made me think more deeply about it. It's good to be able to refine it even further.
> I don't really see how any of that relates to copyright or AI though.
I think the point I was trying to make is that without rules limiting people's behavior, what happens is that those who already have vastly more than other people tend to be able to use that wealth to gain more even faster than the rest.
Being in high decision making ("leadership") positions allows one to capture a disproportional cut of the income. Ownership doesn't even require any work to generate more wealth for the owner. And now, similarly, huge companies are using their position to repurpose past labor with 100% of the wealth thusly generated going to them instead of those who performed that labor.
The high-power unit had 300 grams of Pu-238 in 1965. Given its 87.7 years half-life, only 187g of Pu-238 remaining. It's very hard to do much damage with this amount of radioactive material.
U-234 is ~3000x less radioactive than Pu-238, so having ~120g of U-234 is negligible.
I really fail to see a problem with these tiny amounts of non-brittle material embedded into a solid case. It's still very dangerous, but it's locally dangerous (meters away), not at the scale of whole countries.
I did try using tmux in VS Code's terminal, but still managed to hit this issue with very large scroll buffers. Bumping tmux's internal scrollback buffer capacity to a 50-100k would delay the issue, but it would still eventually occur. In the last few CC versions they did get rid of the old buffer after e.g. conversation compaction, so that's maybe why you don't experience the issue anymore. An Anthropic engineer had even done a PR on tmux's GitHub repo to implement the OSC10/11 shennanigans, so jt seems tmux is not immune to it at all.
Yeah, this company (GeneralistAI) is, in my opinion, the most advanced robotics+AI company in the world. Slightly behind them Google DeepMind Robotics and Physical Intelligence, and then the rest.
In the meantime, China is constructing nuclear cargo ships ([1], [2]) that will be able to transport 14,000 containers at full throttle (200MW) without a need to refuel for years.
Obviously, it's still not done, and yet to prove to be profitable, but their reactor design does suggest that they have a chance to make it work and replace a lot of CO2 emissions.
Nuclear ship propulsion is not a new technology so I have no doubt they will succeed. But where this will and all previous attempts have (see: N.S. Savannah) failed is that ports do not want to accept such ships, because of environmental and insurance issues.
As much as we want the next generation ships with less CO2 emissions, that nuclear cargo ship is not coming anytime soon. It will use a Thorium melting salt reactor, which although not new, is still in experiment. US first proposed melting salt reactor in the 1940s and had operated one briefly in the 1960s. China is building and testing an experimental Thorium reactor. It made the news earlier this month for been successfully converting Thorium to Uranium, perhaps means it is actually burning Thorium as compare to rely on Uranium in previous melting salt reactors. The news also states the plan has 3 stages: experimental reactor, research reactor, and demonstration reactor. The current one is at first stage. The third stage, demonstrator reactor, is scheduled to build in 2035. Therefor that nuclear cargo ship is at least another 10 years from now.
Aren't all military submarines and aircraft carriers using nuclear power already? No reason they couldn't make it work as long as regulations allow them in ports.
No reason the ships can't be built but turning a profit is another thing. US nuclear-powered aircraft carriers cost ~$10 billion, though of course they have many other requirements.
Flying to the Moon can be made to work - it's been done for 56 years. But turning a profit is trickier.
How ironic, a huge "EMBRACING NET-ZERO FUTURE" logo on the behemoth filled with cheap plastic crap that will chill out in landfills for generations to come.
I refer to the RDNA4 instruction set manual ([1]), page 90, Table 41. WMMA Instructions.
They support FP8/BF8 with F32 accumulate and also IU4 with I32 accumulate. The max matrix size is 16x16. For comparison, NVIDIA Blackwell GB200 supports matrices up to 256x32 for FP8 and 256x96 for NVFP4.
This matters for overall throughput, as feeding a bigger matrix unit is actually cheaper in terms of memory bandwidth, as the number of FLOPs grows O(n^2) when increasing the size of a systolic array, while the number of inputs/outputs as O(n).
It's misleading to compare a desktop GPU against a data center GPU on these metrics. Blackwell data center tenor cores are different from Blackwell consumer tensor cores, and same for the AMD side.
Also, the size of the native / atomic matrix fragment size isn't relevant for memory bandwidth because you can always build larger matrices out of multiple fragments in the register file. A single matrix fragment is read from memory once and used in multiple matmul instructions, which has the same effect on memory bandwidth as using a single larger matmul instruction.
> This sounds like a happy ending for the employees of Windsurf and a good deal for Cognition
The employees were robbed from having a big cash exit. Illiquid stock options from Windsurf were converted to illiquid stock options of Devin.
What's worse is that the well is now poisoned. I would advise against joining startups from now on, because I think that there's no upside for employees anymore.
To a degree, this is what copyright was supposed to do.