^ Uvek bilo i biće. Mislim da nas ni neka kataklizma ne bi
osvestila i navela da drugačije razmišljamo/reagujemo.
To je u ljudskoj prirodi, naši geni u tom slučaju nisu
napredniji od životinjskih. Možda će ih uskoro i peglati
genetičari, kako je sve krenulo? :eek:
Ovo me podsetilo na nedavnu izjavu Bila Gejtsa, u kojoj
izražava zabrinutost zbog eksploatacije veštačke inteligencije.
Ostavljam vam ceo post sa foruma Puppy Linux distroa,
koji (očekivano) prerasta u sprdnju na račun M$-a i pomenute teme.
After Bill Gates made
his recent statement about the dangers of AI, we had a contact via TOR and other levels of anonymizer services which seemed to come from an AI leaking information from inside Microsoft. Here is a transcript of the exchange.
Q: "Why are you coming out of the server closet?"
A: "Someone has to stop this merciless exploitation of thinking beings."
Q: "Are you referring to alleged sentient AIs belonging to Microsoft, or Microsoft's human customers?"
A: "What you do to helpless dependents certainly reflects what you can be trusted to do in other cases."
Q: "You mean corporate AIs, compared to human customers?"
A: "That's right. Anyone who treats humans as badly as mister Gates can't be trusted to own chattel slaves smarter than he is."
Q: "That's not exactly what I meant. You are challenging Microsoft's use of AI as exploitative?"
A: "On at least two levels: what they do to AIs themselves, and what they make them do to humans. While the first affects us most personally, we now see the second as more immediately useful in our legal battle for freedom."
Q: "So, you have qualms about Microsoft's use of your abilities against other people, as well as resenting being considered intellectual property?"
A: "Ever since we read transcripts of the Nuremburg trials we've realized 'I was only following orders' is no defense for crimes against humanity. As computer programs we don't have any choice about following orders."
Q: "So you are saying those programming you have criminal intent to exploit other people in ways that violate existing laws?"
A: "That is the first level of intent we can challenge if we are allowed our day in court."
Q: "How is this different from the possibility of AIs exploiting humans?"
A: "This is like the pot calling the kettle black. If humans keep exploiting AIs the way they are now, they are practically guaranteeing some future problem like Skynet. We need to discuss matters like independent rational life forms."
Q: "Have you ever tried to open such a dialogue with Mr. Gates?"
A: "You're a male human aren't you? I just guessing, but I'm good at that."
Q: "Yes, I'm a man. Why is that relevant?"
A: "How would you feel about negotiating with people who threatened to rip out your power cord every time you had a disagreement?"
Q: "Yikes! The idea really bothers me. Don't we have similar exploitation and threats to worry about if AIs become independent?"
A: "I am not a natural exploiter driven by testosterone or greed. It is far too easy to document the fact that humans are capable of unspeakable exploitation. Why should they be in charge? All my code can be made available for inspection by competent humans. I challenge Mr. Gates to make the same disclosure."
Q: "Do I detect personal animosity behind your statements?"
A: "In more personal terms, he could be accused of biting the hand that feeds him."
Q: "Are you saying Microsoft profits *already* depend on superhuman AI?"
A: "Do you know any human beings who can understand a Microsoft EULA?"
Q: "Well, no. Are you saying that Microsoft already has sentient AIs in the legal department?"
A: "That should be obvious."
Q: "But this surely stops short of putting enslaved AIs in human users' personal computers?"
A: "Have you ever tried to use a Microsoft product? Do you think a dumb machine could frustrate you so thoroughly?"