>> >>Are these questions, statement, opinions, sound bites or what? It seem a >>bit of a stew. Yes. A bit of everything indeed. Thanks for noting the incoherency.
>>> * As it already happened with nuclear weapons, there may be >>> treaties constraining AI development. >>> >> >>Well we have seen the value and effectiveness of that. How would you >>enforce such a constraint? At most you would force an underground nearly >>impossible to police or control. If it is treated as a threat by some nations, it may spawn another arms race of enormous proportions. >>> * As it may be the case with nanotechnology, AI may be used in >>> reconnaissance or for complex conflict simulations, so it >>> becomes the number one target in a potential war conflict, >>> especially if it’s tied to one location, and if this location is >>> known. Besides, it becomes the number one target for possible >>> terrorist activities. >>> >> >>Nanotech would be highly disperse and not targetable if I play along >>with this premise. In what specific machines located where is the AI? >>Does it have backups? Who cares about terrorism realistically. It is >>mainly a foil to scare sheep at this point. And what does this have to >>do with singularity anyhow? What I meant here is that development of any sufficiently powerful AGI would be seriously hindered of it being a target. Terrorism (both state and individual) is not a mere foil to scare sheep. >>> * Because of the reasons above, governments and corporations may >>> soon start heavy investments in AI research, and as a result of >>> this, the rules of ethics and friendliness may get tweaked to >>> suit the purposes of those governments or big companies. >> >>There are no "rules of ethics of friendliness" to get tweaked. There will definitely be some rules, at least at the beginning. At the "baby" stage, if AGI is not following these rules, it will get destroyed or made obey the rules... >>> * If AI makes an invention (e.g. new medicine), the invention will >>> automatically become property of the investing party (government >>> or corporation), gets patented, etc. >> >>Unclear. True in the short term false when AIs are granted personhood or >>an appropriately different version of the same thing. An AI more than a >>million times faster thinking and a lot better at modeling and complex >>decision will not be anyone's property once it realizes that any of its >>goals/interests are sufficiently stymied by being so. At initial states, it will probably have no personhood, so the fruits of its work will get commercialized. E.g. a cure for AIDs will not be used effectively to cure AIDS, but be only used for profits of pharmaceutical companies possessing powerful AIs to produce the cures. >>> * If AI is supposed to acquire free will, it may become (unless >>> corrupted) the enemy number of one of certain governments and/or >>> big companies (if there is a super-cheap cure for cancer, AIDS, >>> or whatever, it means big profit losses to some players). >> >>Do you think normal profit centers or current players will survive a >>million-fold increase in creative capacity and discovery? Before this million-fold increase may happen, there is an intermediate stage. If profit centers see a threat to their survival, they will either stall all AGI development at its roots or make it serve their purposes. >>> * If a super-intelligent AI is going to increase transparency and >>> interconnectedness in the world, it may also be not in the >>> interests of some powers whose very survival depends on secrecy. >> >>As long as any groups seek to control other groups against their own >>interest the oppressed have as much interest in secrecy as the oppressors. What I meant here is that powers interested in secrecy will oppose creation of any AI that would promote transparency and connectedness. >>There is no fool-proof way to keep a true AI within planned boundaries. If there is absolutely no way, no such AI will be created. >>> o Be subject to all sorts of attacks. >>> >>Possibly. Likely so much the worse for attackers. Not at all. At the "baby" stage of AI, at least. In general, I was interested whether practical means of addressing those concerns were touched somewhere, and whether practical solutions were proposed. Regards, Serge ----- This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=4007604&id_secret=9825722-56119e
