One must “evaluate the actions” in order to ration them out. That is the only purpose in evaluating any action. Being able to update with oncoming new information is no magic trick either. That is what sensors are for. Intellacars are doing that right now in California. Autopilots mechanisms have been doing that for decades. Rewriting algorithms on the fly is far more sophisticated, but is also happen in many systems. Choosing the better way to store or restore altered data is a common AI function. Patterns of common use and anticipated use play a role in choosing which way to store data. Internet systems are doing it every day.
Just to give you a clue. I the early 1980’s, I wrote a small operating system for diagnosing and trouble-shooting a telephone system. The intent was to allow a technician to electronically “play with” any part of the hardware so that he could isolate and monitor the telephone system while it was still in use. So the small operating system had to also have a language with which to communicate with the technician. And as the technician told the diagnostic system what he wanted done, the program would figure out which lower level assembly language codes to use in order to assemble a program to meet the request, then lay that program into place, and run it as per the technicians orders.
The point is that my little operating system and communication program was writing and implementing code for the technician so that he wouldn’t have to learn the variety of languages potentially involved 30 years ago. The system was “hardware portable”. And in an automatic mode, it would search through all of the possible ports and memory, testing for any kind of anomaly. It could identify anomalies and ask the technician what he wanted to do about it, what kind of image that he wanted to see on his oscilloscope.
Today, things are far, far beyond that little device. There is no decision that you or anyone can make, by any means, that even I couldn’t have written a program to do instead and automatically go on to the next, no matter how complicated you felt the decision was. Every single thing you do can be modeled and automatically emulated. Machines today can even do the modeling and emulation code writing themselves. EVERYTHING can be automated, even learning new things to automate (even today).
The software world doesn’t actually have to invent anything new. They like to at times. It’s nice to get things a little faster and/or smaller, but almost every new item is merely a rearrangement of already documented items, in a new package to use or sell in a new market.
They are not inventing androids. They are assembling them to do different things in different ways. They are experimenting with which way they want. It is not an issue of having to invent anything.
That is absolutely right. Actually, softwares are not true inventions, as we understand that word. Those are merely more and more complicated presentations of what was invented once and that was Binary language. We can put that into the categoty of invention, if we like.
New softwares are merely the repetition of that invention in new verticals.
Like the concept of combusion engines. one can use it in a bike. One can use its bigger version in the car, even bigger version in the tank too. But, the engine of an tank is not a ture invention. It is the same concept that was used in the bike.
The same is with the softwares. Unless their basic binary concept does not change, it is all the same except complexity.
Neither being a soft or hard ware specialist, in any sense of the word, there probably is no flaw in Your presentation, James. However, the point is, and this THE critical point in any system, whether be it human or machine-like, is that the stage has not passed, where, machines are 'taking over' Now i feel i am letting You down, with an over generalization as a response to technical analysis, but, with a goal of UNIFICATION, of de-differentiating the technical with the logical, one should have bearing on the other.
S0 that to imply that the matter of absolute control is still sustained by the programmer over the program, i feel is still a valid assumption, since if it were otherwise, machines would have ways to punish by override any attempt at an intrusive control. If such was the case, the system would most definitely be shut down, or at least ,modified to the extent, that such an unlikely event could be prevented.
2004: The Space Odd essy, is still at this point wildly futuristic, in my opinion, but then i am only trying to make philosophical sense.
Obe, in a war where machines are fighting machines (already begun throughout the world) would you program your machines to wait for each decision to be made by you when the enemy is allowing their machines to much more quickly make their own decisions?
See I just don’t get it. Machines like humans have Achilles heels. Programs in a machine can be stopped. A virus soup can be sent to any machine that is programmed to make decisions. Just hack the darn thing with viruses it will stop it or slow it down or turn it. You would have to be very careful about what viral commands are sent. A barrage of such would eventually work.
You see, this is why humanity has so little hope. You get your sense of reality and hope from TV and Hollywood films. Do you seriously believe that the military designs combat androids and weapons to be such that the enemy merely has to “hack into them” and turn them against their creators? Even if they were that ridiculously dumb (after all, we are talking about the military), don’t you think that the military could even more easily hack into them and reverse their infection?
Your PC is so very hackable only because they wanted it to be that way since the beginning and didn’t realize just how clever the populous can be. MACs are not susceptible to such problems. And neither are military combat systems. Communication networks and intelligence systems are not simply inherently hackable.
That point has merit, nevertheless, there probably are programs and programs. In programs where independent, non critical decisions can be made automatically, no oversight is needed. But in decisions involving the relationships between various mutually relevant programs, where their interaction at this level may cause a systemic meltdown, there needs to be oversight. There are decisions, machines simply are not trusted to make, even if the science has gone that far, foe the very reason that decisions made by machines may turn out to be counter-productive.
This is only hypothetical, and the proof is simply that machines have not yet demonstrated any counter-productivity in themselves, except through external hacking.
Uhhuh. Fuck Hollywood crap.
Only complete morons do not put a remote shutdown procedure on such a machine. And only a moron thinks that their work cannot be screwed with . If it is programmed it can be shut down even if there are redundancies.
The point was whether they could be “merely hacked into”. They cannot be “merely hacked into”, especially when they are already programmed to prevent it and are much, much more intelligent than any hacker.
Yes, that's possible, but, who created who? You programed the computer, and unless, Your program consisted with an ability for an autonomous re-program within the program, computer intelligence will be merely a trait of Your own abilities, whereas Your intelligence will be a type which is pre programed by You Yourself (through Your own efforts at education) and by God Almighty, coming through as a genetic 'gift'. The computer owes it's intelligence to Your program, similarly.
LOL show me where I said “merely hacked into”.
A human hacker is just a director for the machine designed to solve the problems. The human just says sic him boy…
As a human, you are already “hacked into” by what you believe to be factual information, which actually isn’t. That is what hacking is all about … getting past the filters and causing illegitimate persuasion (also known as “hypnotism”, the entire priority of the media). They too use “information overload” as a technique.
James, I can’t make these machines that make machines that make machines but, clearly people know how. The ability is there. Did I say it would be easy? Nooo. I would bet though that there is government laboratories already working on it.
You create the tech, you know that someone else is doing the same so you also start a program to stop yours and theirs. Its been in the process for decades. Look at the Ebola outbreak… Certain governments suddenly cough up “experimental” meds for it… A: the stuff they gave was proven to help but, not cure. This tells me they have enough of an actual cure to cure aa specific amount of people.
B: They can stop the outbreak cold but, will not because of power ego.
And they also plan to use the virus as population control.
Now, hon, the shit is out there to stop machines its a done deal. The only question that stands out is: When it will be used.
When it comes to viruses, all of that is a fine argument, and very largely true and easy to imagine. But then you switch suddenly to proclaim that it also applies to machines, "the shit is out there to stop machines its a done deal". What kind of thing, “shit”, are you imagining?
A killing thinking machine built, is like a deadly virus. The people that create a deadly virus will not purposely release that virus unless there is a cure. A thinking machine is quite capable of getting a dangerous glitch due to lack of foresight or damage. A glitch can cause the machine to backfire on its makers. There is not just one person building and thinking of potentials, its a team. A team will have more foresight than one. With this knowledge it is more than safe to know beyond doubt there is a remote kill switch built in. The opposing side knows this and so would have a team finding the programmed kill switch. The cure to the virus.
One sure way to stop is to flood known signals with probable commands and programs. A higher computer near AI status would have the abilities. Now we both know there is at least two major countries with such computers and highly capable teams… The shit.
I see. So tell me, what cure do did they have for the radiation poring out of Japan, Chernobyl, and Iraq that they aren’t telling anyone about?
You seem to have been infected with the hubris notion conflating Man with the all knowing and all wise God of legend… ?
A “team”? Emmm… so there are no competing enemies? No egos within the “team”? A choir of saints? They are just good-ole-boys very altruistically and carefully ensuring the love and harmony of humanity? I didn’t know you were another deranged, delusional liberal socialist.
Except that the first “team” knows that the second team knows and thus plays it differently than that obvious trap. And the second team knows that the first team knows that the second team knows and thus attempts to figure out the alternate trap. It gets really, really complicated, doesn’t it. In fact, it would take a really sophisticated computer to figure out how to play such a game. But then, of course, both “teams” know that and are thus busily making sure that their computers are smarter than the other team’s.
It all kind of makes you wonder what those truly, extremely intelligent computers will decide what to do with such a silly creature as man.
Chernobyl was one example of the very nature of Man, assuming that he has control of the powers he unleashes on the Earth, “we can always just turn it off”. Did he learn? Nope. Fukushima came next and is still grandly destroying life. Has he learned yet? Nope. Recently many attempts have been made to form a black-hole based on the THEORY that it won’t get out of control.
Science wants to excuse itself constantly on the basis of everything being merely a theory and possibly incorrect. And yet science always wants to “play with” the most enormously devastating powers ever conceived, such as black-holes and androids that are not merely alive, but 100 times smarter than any human could ever hope to be.
I am still not convinced that black-holes in space are not merely the remnants of prior “evolved” haphazard life forms (similar to homosapian), playing with their version of “Science”.