I find it slightly sad that when our leaders talk of Technology and Innovation - they often mean these 'tech' companies that essentially work out how to better sell advertising and occasionally provide a useful service alongside this.
Where is the Bell Labs? The Skunk Works?
We have incredible problems facing us such as Climate Change and decarbonisation seems like it will be a very difficult challenge. And yet we focus on banal "innovation" in frivolous things.
I agree completely.
We are already in an age where much menial work can be automated and AI seems to be well on the way to automating a lot of menial information work too. We need to focus on creating a growth mindset and a sense of wonder and curiosity that will serve the children whatever the future may hold - not just creating a holding pen so their parents can go to work.
It's okay, but I would have preferred more of a focus on the actual technical aspects about how Cybernetics worked and helped run the country rather than a focus on the people.
Maybe it becomes like that later, I am up to the third episode.
It's revenue share based on how many streams you get. Big record labels can probably negotiate a better share, but if you sign to one there's no guarantee you will actually see that extra money.
Record labels ripping off music artists was incredibly common in the time before streaming, so I imagine it still is today.
It's revenue share with the record labels. If those labels don't pay the artists well that's a different issue.
Fundamental science is mostly publicly funded though and has little immediate practical application. The lack of funding in much of science also shows the problems this approach has.
Yeah, I guess it depends if the copyright is broad enough to offer protection while not becoming too broad and stopping innovation.
Yeah, but some things cost a lot of money to develop. The higher the cost of the R&D, the less likely it is to occur without some patent system. Although I agree that in programming specifically the Open Source model seems to work quite well - look at the Apache Foundation.
You could have a model where all research was done by a public body or something like the Apache Foundation, but this reduces innovation as it means there is less opportunity for some people to try something that may not be considered likely to be successful, as publicly funded research tends to focus on the safest path. For an example, look at how public nuclear fusion research is continuing on the traditional toroidal tokamak model with ITER compared with the more experimental designs being tested by private companies such as Helion, Focus Fusion, Tokamak Energy (they are using a high aspect-ratio 'spherical tokamak).
Yeah, I agree on those examples. They should be able to patent their particular implementation - like maybe it took a lot of R&D to work out how to get server response times fast enough for one-click to work, or to get loading times fast enough to have a mini-game in the loading screen etc.
But they shouldn't be able to patent the entire concept. That's ridiculous.
Yeah, I think with software due to the low barrier to entry etc. it makes sense for it to be further towards the less protections end of the spectrum.
But still, if you'd paid a load of PhDs to come up with some really clever algorithm (think of like how Shazam had it's music recognition algorithm long, long before modern ML) and then someone could just steal it well, it'd harm innovation and ultimately the tech industry and investment would go elsewhere and those clever PhD grads just wouldn't find employment.
It's a balance that depends on the properties of each industry, but I don't think that no protections whatsoever is ever a good answer.
Has it ever been good at mathematical/logical problems? It seems it's good at text-based problems like imitating a writing style or even writing code, but if you ask it a logic puzzle like "if two cars take 3 hours to reach NYC, how long will 5 cars take?" it often fails completely.
Humans are capable of both understanding language and logical thought, I'm not sure if the latter will ever be easy for the LLMs to do, and perhaps older Symbolic approaches to AI might perform better in this space.
I don't have anything against trains, but our rail network is really limited.
If I want to go from Barcelona to Madrid, it's easy and actually more convenient than flying albeit more expensive.
But if I want to take my kids to go and see the cool medieval castle in the mountains? There's no train going anywhere near there.