4 ms·
> So I'm down to my final two guesses. Either the guy thought he didn't do anything wrong (really? taking confidential documents and hardware??). Or the article
by antpls 8y ago
> So I'm down to my final two guesses. Either the guy thought he didn't do anything wrong (really? taking confidential documents and hardware??). Or the article heavily distorts the facts.
Probably a bit of both. And also many other facts we don't know. Anyway, Apple only should be blamed, imho. If people efforts and project momentum are the real plus value, then data alone won't be sufficient to steal a technology.
It shows that international cooperation about global-impacting technologies should be the norm, rather than a race to monopolize a market.
- fauigerzigerk 8y ago>Anyway, Apple only should be blamed, imho. If people efforts and project momentum are the real plus value, then data alone won't be sufficient to steal a technology. That's a weird argument. No one said it's sufficient, but "stealing" data certainly helps. Also, what do you mean by "people efforts"? The only way to prevent every single employee with access to trade secrets from leaving is to make workers corporate property and lock them up on a remote island. Surely, that's not what you're suggesting, is it?
- antpls 8y ago> That's a weird argument. No one said it's sufficient, but "stealing" data certainly helps. IF the employee wasn't aware that he/she was stealing, maybe he was being told internally what he/she was doing was OK. Maybe someone suggested him to do so, only then to be accused after. Anyways, the ability for an individual to transfer so much knowledge should remind you the MP3 and p2p era. The patents and secrecy of big corporations makes NO SENSE in a world where communication is instant and easy. Not better than China censorship.
- fauigerzigerk 8y agoSo what does make sense if the problem is how to protect the value of investments into the acquisition/generation of knowledge?
- antpls 8y agoIt makes sense to replicate and sharing knowledge among as many people and organizations as possible. Replication is the best way to protect knowledge. The old model is to see knowledge as words written down on a piece of paper, and only knew by a few people. It does not work in a modern world where knowledge can be shared easily, sent at speed of light and replicated on many support. I don't know about you, but I don't have millions of dollars sitting in my bank account, I prolly don't care about previous investments and decisions made in a world I didn't exist, but I care about living a good life and living as long as possible. The investments should change direction.
- fauigerzigerk 8y agoLet me rephrase my question: How do you suggest a profit driven company can justify employing people to create new knowledge if that knowledge and all its derivatives and representations are immediately available to everyone else without any restrictions? I am very skeptical of patents as a matter of principle, because banning people from using their own original ideas just because someone else happened to have a similar idea seems extremely unfair and prone to misuse. I am also skeptical of putting copyright interests above all else (although I'm not against copyrights in principle), going as far as building a global surveillance state and imposing draconian penalties in order to enforce extremist copyright laws without fail. I am also skeptical of very restrictive no-compete clauses in employment contracts or any sort of anti-poaching activity by corporations. And it's exactly _because_ I am skeptical of all the above approaches to protecting investment into knowledge generation that I'm very hesitant to also oppose keeping trade secrets. Creating new knowledge has to have some direct economic benefit above and beyond the general benefits of human progress (which is also important of course).
- antpls 8y ago> Let me rephrase my question: How do you suggest a profit driven company can justify employing people to create new knowledge if that knowledge and all its derivatives and representations are immediately available to everyone else without any restrictions? I have two arguments : 1) The process of creating new knowledge alone won't be valuable once we have enough processing power to evaluate scenarii faster than a coordinated human effort. With vulgar words : AI will make the creation of knowledge a commodity, and my personal opinion is that it will happen before the next human generation, unless some people actively work against that idea to make sure it doesn't happen (because they try to defend previous investments for example.) 2) Knowledge alone is not enough to accomplish or execute anything in the real world. Actually building car requires factories, materials and logistics. Of course, if we have an AI solving all sort of problems, we may ask it to find an efficient way to build car. If we continue the reasoning, we enter a world where the meaning of an human life is completely different than today, with completely different forces acting on it. We may don't even need the cars anymore. Edit : I mean, the argument of "it gives job to people" won't stand forever. We will have to seriously re-think the system at some point.
- pertymcpert 8y agoThat’s a completely ridiculous line of reasoning. How can you honestly write that it’s no better than censorship. Have you been spent any significant time living behind the great firewall?
- antpls 8y agoBlocking incoming information or outgoing information requires a firewall either way. Here the firewall is the patent system, or any cultural belief that trade secrets create more value than open systems.
- pertymcpert 8y agoThat's not what I asked.
- antpls 8y agoIf you can't read it between the lines : Yes, we are all living behind some sort of firewalls, either technical, legal or cultural.
- pertymcpert 8y agoYou said "Not better than China censorship." That's ridiculous. There are degrees, shades of grey.