Didn't they say that they launched an effort to evaluate their model on all open Millennium Prize problems, and then narrowed down to Navier-Stokes as the most promising after seeing results on simplified versions of the problems?
Yes, you're right. It seems like they claim to have started broadly and narrowed it down based on some progress. That leads to different questions but does answer my initial "fishy" point.
Google also spent heavily setting up deals with other platform owners, invested in Chrome and Android to establish Google search as the default option which most users accepted. If integrations of good enough AI features are made within existing platforms most users will probably accept using those and not think too much about whose model is powering it.
People can also ask questions that it is unreasonable to expect people to know on the spot too and after a while the role of the questioner in the meeting can be questioned. :)
Includes an example of a Flock camera flagging a number plate that the police pulled over, a family car SUV believing it to be stolen, but the stolen vehicle was actually a motorcycle with the same plate number but a different state!
Sure, you don't pay millions of dollars in engineering for a permit you obviously aren't going to be able to get. That's a big part of the problem, they're unpredictable.
Much like his earlier work Westworld which was also scarily prescient for modern times.
> These are highly complicated pieces of equipment. Almost as complicated as living organisms. In some cases, they have been designed by other computers. We don't know exactly how they work.
or The Andromeda Strain. I was on the edge of my seat for the last half of the movie as a 10 year old when it came out. The book still holds up as a medical mystery - thriller and warning of unintended consequences.
The competence threshold can vary according to the model's complexity and cost to run, but the broader point remains that having a human operator of greater competence that can evaluate the output and understand potential harms that might result seems to be a valid one. Depending on domain, it can be irresponsible to deploy systems you don't fully understand that could harm others. Responsibly overseeing 37k lines of code a day sounds exhausting.
Not as many changes to the files under library as I expected to see. Most changes seemed to be under a single ‘add stuff’ commit. If some of the solvers are randomised, then repeatedly running and recording best solution found will continually improve over time and give the illusion of the agent making algorithmic advancements, won’t it?
I guess my point was that I don't see many algo changes in the commit history, which is a shame if this has been lost; library/* files are largely unchanged from the initial commits. But each time the agent runs, it has access to the best solutions found so far and can start from there, often using randomisation, which the agent claims helps it escape local minima e.g. 'simulated annealing as a universal improver'. It would be nice to see how its learnt knowledge performs when applied to unseen problems in a restricted timeframe.
Locking down children’s devices doesn’t stop adults sharing illegal content with other adults though, so there would still be pressure to monitor communications between adults.
At some points, laws become an ineffective tool to prevent malevolent people to act in detrimental manners, no matter what it states. But prejudices of wicked states will always continue to impact more badly general public as ever more drastic laws lacking any balance become enacted.
reply