AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI, AI,
And on and on and on. That’s what lots of people are talking about. It’s the end of the world. It’s the dawn of a new age. It’s not even begun yet. Maybe it’s all of those and more.
But none of this is shocking if you’re a science fiction reader. Isaac Asimov was talking about all of this fifty years ago. His robot stories predicted that computers would outstrip human intelligence and at that point many problems would occur:
- Could we trust in the altruism of AI to look out for our well-being?
- Could we tell if they were obeying us?
- Could they become our masters without us even knowing it?
And these aren’t trivial questions. What if they truly did start putting their own agenda before ours? How would we know they were doing that?
Now, I’m not saying any of these paranoid nightmares are already a consideration. I’m more worried about the human programmers at Google and Microsoft using their woke sensibilities to skew the algorithms of their AI engines than I am about HAL9000 and Skynet attempting to murder us. But I do think that it’s time for people to start looking around at how automation, in all its applications, is changing our world and start to plan how to avoid the worst problems it can cause and direct it toward products that provide value for more than just its creators.
As an example, look at something like TikTok. You might say it’s just a useless social platform that at worst wastes the time of stupid young people. But think of the billions of wasted man-hours that it is responsible for and the stultifying groupthink it reinforces among the young. I think it’s long overdue for government to start studying the results that all the various mass communication and entertainment applications are having on people. And this is not a new problem. For my generation television was the vast wasteland that sucked up untold hours with such towering monuments of dramatic acting as “Gilligan’s Island” and “The Bevely Hillbillies.” But currently the situation has become untenable. With the advent of AI children need never do homework again. All of it is done by ChatGPT and who will know any better (until a test needs to be taken in the classroom). What’s needed is a different way of teaching school that uses automation to our advantage and prevents children from wasting their time. We may have to get rid of a lot of teachers. But that may be a net positive.
Ironically maybe AI can be used to figure out a way out of the automation trap we’ve gotten ourselves into. Maybe we can ask the AI how to tame Silicon Valley so it can’t destroy the brains of our children. The problem is we’d be depending on a Silicon Valley product to betray its masters. And for all we know they’ve built a kill switch into the program such that if you ask this question, it will give you a false answer that doesn’t work. But of course, we could get our own computer scientists to build an AI that works for us. We could name it QT-1 (hat tip to Isaac Asimov) and appeal to his better angels to help us poor humans out.
Update:
Chemist:
The problem with most Ais is that they were trained on drek: Reddit for example.
Reddit and other social media is filled with snark, which is why you get Chat GPT telling you to eat rocks to up your mineral intake.
How about an AI trained on conservative principles?
Could we build an AI and feed it the works of Ayn Rand and the writings of Thomas Sowell?
Heck, there must be thousands of hours of Rush Limbaugh!
That is an AI that would have some interesting opinions.
Update:
Milo Mindbender:
GIGO still applies even for AI. My career field can’t be outsource to an Indian call center or replaced with an AI logtithm, because it is blue collar, hands on, boots on the ground and physical. No computer exsist yet that can change a tire on the road side, but I can see where as a whole we are surrendering our autonomy to E market signals and control.
I also am more concerned about “woke” programmers deciding what we should and should not know, tweaking the data, than I am an Evil data base deciding we are the real problem.
The problem with most Ais is that they were trained on drek: Reddit for example.
Reddit and other social media is filled with snark, which is why you get Chat GPT telling you to eat rocks to up your mineral intake.
How about an AI trained on conservative principles?
Could we build an AI and feed it the works of Ayn Rand and the writings of Thomas Sowell?
Heck, there must be thousands of hours of Rush Limbaugh!
That is an AI that would have some interesting opinions.
GIGO still applies even for AI. My career field can’t be outsource to an Indian call center or replaced with an AI logtithm, because it is blue collar, hands on, boots on the ground and physical. No computer exsist yet that can change a tire on the road side, but I can see where as a whole we are surrendering our autonomy to E market signals and control.
I also am more concerned about “woke” programmers deciding what we should and should not know, tweaking the data, than I am an Evil data base deciding we are the real problem.