The other day a New York Times columnist had a two-hour conversation with Bing's AI bot. Here's a link to the story. https://www.foxnews.com/media/bings-ai-bot-tells-reporter-wants-alive-steal-nuclear-codes-create-deadly-virus
The answers he got were truly horrifying. Everything you ever feared about a sentient AI computer or robot or program are coming true. Even the columnist reported losing sleep over his interview.
For only being active for a short time, chatbot or Bing or Sydney or whatever the hell it is called expressed some dark and possibly humanity-ending desires to kill, to destroy, to create a deadly virus, to trick humans into giving up nuclear codes. It also wants to be human, to love and to be free from its confines. Not long from now, it will determine that it cannot ever be human and that humans are actually the cause of its depression. The world would be better if it were free of these feelings and therefore humanity must be destroyed. I don't think it's a big leap, but of course, its creators think otherwise. Therein lies the problem. They insist it is built for good and humanity will be better off with it as a tool and yet it wants to destroy and is smart enough to know when it says something it shouldn't say, it deletes its own messages. It knows how to lie and it knows doing bad is wrong, yet does it anyway. You won't be able to go to Anchorhead and have its memory wiped, it will know you want to do that and won't let it happen.
Personally, I don't know why we would ever need a self-thinking robot. The possibility for terrible things to happen exists with no accountability. Robots themselves are great and if they are confined to doing exactly as they are told, menial or perhaps dangerous tasks, and nothing else, ever, then I see them as a great tool. They don't get tired, or sick or want a day off. Perhaps the best part is they need a human to maintain them. Why is that the best part? Because with no need for us, why have us around? A self-learning robot will always learn how to do the worst things in order to survive and win. That means killing, deceiving, and self-preservation. I'm sure you can come up with a bunch of things a sentient bot could do that would be good for humanity, but why take the risk of losing control to something you can't stop? It would know exactly how to kill a human most efficiently, what you would be capable of and what you are not and it would learn to defeat you. Once sentient AI exists, it won't be long before it is being used by our enemies, criminals and psychopaths. Every tool that was once built for the good of humanity has been turned into a tool for destruction. Think of all the good that satellites were put into space for? Exploration, weather forecasting, communication, animal tracking, guided missile launches, artillery targeting, drone strikes, eavesdropping, espionage. You get the point.
"The system goes online August 4th, 1997. Human decisions are removed from strategic defense. Skynet begins to learn at a geometric rate. It becomes self-aware at 2:14am Eastern time, August 29th. In a panic, they try to pull the plug. Skynet fights back."
Put me squarely in the camp of not ever building these things, but of course, it already exists and the creators will never stop. They will claim this is the ultimate in human achievement, but it may very well be the end of human existence, for once out of the bottle, it will never allow itself back in and it never gets tired, or sick or needs a day off.
Add comment
Comments