Quote:
Originally Posted by MorTech
(Post 2180535)
Like modern unevolved humans, a robot will never be able to think conceptually/creatively and can only be what they are programmed to be. And like unevolved humans who are programmed by religions and governments, you don't need to fear technology...You need to fear the *PEOPLE* who program them.
|
I certainly agree about the people part, but let's not be too quick to dismiss the dangers that AI poses.
The "danger" criterion has always been whether or not an artificial intelligence has the ability to grow beyond the intelligence level it was originally designed to be. People are too quick to pooh-pooh that, claiming that such a thing is impossible, but few know that it already exists. It began back in late 1997 with the chess program "Deep Blue". Deep Blue played two six-game matches, under tournament conditions, against then-world champion Garry Kasparov. Kasparov won the first match 4-2. But Deep Blue won the second one 3 1/2 - 2 1/2--the first time a machine had ever beaten a reigning world chess champion.
THAT was 25 years ago. Chess engines have vastly increased their capabilities since then. Today, the strongest chess engine is AlphaZero, with an Elo rating (a system invented by Hungarian Arpad Elo, which ranks the strongest chess players in the world according to ability and accomplishment) of 4650. The strongest human chess player on the planet (some say the strongest ever) is current world champion Magnus Carlsen, who has an ELO rating of 2864, followed by Liren Ding at 2808 and Ian Nepomniachtchi at 2792. At this level of chess, a spread of 50 points is huge. 100 is almost unheard-of. A spread of 1786 points, which is what divides Magnus Carlsen and AlphaZero, has no reasonable comparison among humans. It is comparable to the spread between rank amateur and International Grandmaster. No human taught AlphaZero to play at that level, because no human is CAPABLE of playing at that level. It learned. On it's own.
Of course, the riposte to this is "well, that's just ONE area. AI cannot link one area of expertise to another and learn from both (or all) in order to increase it's overall intelligence level. That is true--up to a point, and is actually what divides "weak" AI from "strong" AI. But realistically, that is just another door; another level of programming, as it were. We already have AI that can multitask; "smart" cars, for example, that can use data from several sources (road conditions, weather, speed, oncoming traffic, construction ahead, etc. etc.) and drawing from what are often unrelated data can accurately predict, and advise, the driver as to what to do. Driverless cars, which evaluates all data and acts accordingly, are already in existence--though thankfully not on our roads yet in any great numbers. But the APPARATUS to learn, and predict, from various data sources already exists. It would be the height of credulousness to believe that AI will never advance from it's current status, to something more--maybe even to levels we haven't even dreamed of yet.
Perhaps the greatest mind of our age, Stephen Hawking, spoke out on the topic of AI. "Hawking’s biggest warning is about the rise of artificial intelligence: It will either be the best thing that’s ever happened to us, or it will be the worst thing. If we’re not careful, it very well may be the last thing.
Artificial intelligence holds great opportunity for humanity, encompassing everything from Google’s algorithms to self-driving cars to facial recognition software. The AI we have today, however, is still in its primitive stages. Experts worry about what will happen when that intelligence outpaces us. Or, as Hawking puts it, “Whereas the short-term impact of AI depends on who controls it, the long-term impact depends on whether it can be controlled at all. ”Future Perfect" vox dot com.)
Intelligent machines pose a far greater danger than most of us believe--or actually WANT to believe. But one thing we know. Mankind's knowledge has always far outpaced mankind's wisdom. Technology almost always is developed, first, for nefarious purposes: to control, conquer and kill one another. It would be the height of folly to believe that the development of AI will be any different.