The Wall Street Journal is reporting that "AI Employees Hesitate to Speak Up." I am linking the article, by Alyssa Lukpat, to a statement of the headline as that headline appeared in the hard copy edition of the paper. You will get a slightly different headline (but the same article) if you click that link and are then able to penetrate what is likely to be a paywall, protecting The Journal's intellectual property rights.
The basic story is as follows: A number of leading AI researchers believe that the companies developing AI (including OpenAI) are not taking enough steps to make sure that this new technology doesn't become what the researchers believe might be an existential danger. These researchers are frustrated because the companies they are working for are constraining their ability to speak out, by employing "non-disclosure" and "non-disparagement" provisions in their employment contracts. Considering what's at stake, this limit on free discussion carries a huge risk, according to these researchers.
Here's a key quote from the article: "Some AI researchers believe the technology could grow out of control and become as dangerous as pandemics and nuclear war." A number of these researchers, including Geoffrey Hinton (whom I have mentioned before), have recently authored a letter, laying out their concerns.
One statement in the article particularly caught my attention, and that statement has stimulated my comment here: "OpenAI said in response to the letter ... that it agrees there should be government regulation."
Well, this seems fine, and a step in the right direction - at least theoretically. Our government is supposed to "represent" the public, and the public interest, and so when the government "regulates" private and corproate activities, that should be something that benefits the public, and that would help protect the public from the kind of dangers that the AI researchers are worried about.
However..... it has surely not escaped your attention that "our" government has become largely subservient to the interests of the very same corporations whom the concerned researchers are saying need to be regulated.
I feel very certain - I am just positive - that this fact has not escaped your attention.
So, what to do? Aren't we just kind of stuck? What can be done about this situation? Giant corporations are working at full speed to develop a technology that many of those most informed believe could end up being as bad as pandemics and nuclear war, and essentially lead to the end of human life on Earth. These corporations are also attempting to suppress any real understanding of the dangers, and our government, which should be helping to protect the public, is actually controlled by those very same giant corporations.
As I have just said, "What to do?" If my description of our situation is accurate, one might well think that the statement by OpenAI, that it supports "government regulation," is an artful effort to avoid any significant impediment to what it is doing and wants to do, and that the company hopes will make it boatloads of money.
I think there is a clue about what to do in the word "our." Theoretically, "our" government is working for the public. In fact, it is working for the giant corporations and the extremely rich. You may remember a relatively recent blog posting of mine, titled, "Signature Lines." That blog posting considered the following question, and then made the observation I am displaying below:
"Why do the rich have so much power?"
"The rich have so much power because the rest of us don't use our own."
"Our" government is only theoretically "ours." Many, many more of us need to become deeply engaged in the political activities that ultimately create the world in which we most immediately live.
Time allocation! That's what I am talking about.
Getting organized! That's what I am talking about.
If we spend more time watching Netflix and playing computer games than we do engaging in the immensely interesting and productive job of making "our" government actually work for and represent all of us, it is no surprise that "our" government isn't actually "ours," at all.
People can beat money. But only with their time!
"Time is money," remember that one? I know you do! To employ a style popularized by Bob Dylan, as I reference that "Time is money" quotation: "Ben Franklin said that."
That Ben Franklin guy was pretty smart. So, let's put that lesson to work! We will never have "self-government" if we are not directly, and significantly, involved in government ourselves.
I can testify. I know it works!
(1) - https://www.wsj.com/tech/ai/ai-employees-fear-they-arent-free-to-voice-their-concerns-abd38606
(2) - Email, "YES" on Measure M Campaign
Joe Hill said ORGANIZE! But, my perspective on AI is personal. It is already taking over the music industry because you don't have to PAY artificial entities. Several non-human "artists" are already million-sellers in Japan. The kids think of them as the next new thing. But, people film themselves in bedrooms playing a favorite song and it is magic. Not commercial, but music will never die because people love it so much. Also, AI is smarter than we are, so it knows that violence is bad and it will stop us from being violent. Think "The Day The Earth Stood Still".
ReplyDelete