463 Comments
Nov 22, 2023·edited Nov 23, 2023Liked by Robert Reich

The government setting regulations is the only thing with any hope, but it will likely fail for two reasons. One, we know that Congress is susceptible to corruption either in the pursuit of money or power or re-election. Two, the vast majority of them are technically incompetent to create any rules governing it and unwilling to listen to those that could do it.

Expand full comment

It's also a world wide problem. Life as science fiction. Open AI is just one company.

If nations are working on matters like nuclear research, space travel, genetics, also AI!. At one time it was suggested that DNA can be used instead of chips in computers. https://www.sciencedaily.com/releases/2021/09/210914184825.htm

Imagine man as automaton. Man as android. One mad scientist in a lab at NIH......controls the universe.

Imagine Putin's control of the algorithm for world domination. Kim Jong Un. The Ayatollah Ali Khamenei. Bishop Strickland. Franklin Graham.. MAGAT Mike.

Expand full comment

And let's not forget the treasonous, felonious mango madman. Although he's probably not well versed about the numerous capabilities of AI, there's a very strong argument why he shouldn't be anywhere near it.

Expand full comment

Thankfully, Trump hasn't found a way yet to make AI work for him personally. And the US absolutely needs to finally put him in the safe place he belongs. Else there will be other problems that then will need immediate solution, there won't be the time to occupy with AI.

Expand full comment

Vau Geha: Yes, tfg thrives on confusion and chaos.

Expand full comment

Mango madman..that is a first, love it..

Expand full comment

Who specifically is your opinion of ?

Expand full comment

Rick Calegari : Or any of his 'clever' 'henchmen'.

Expand full comment

True!

Expand full comment

From the time I was in grade school, I thought the world would be destroyed with nuclear weapons. I didn't think I would grow up. Now I'm a great-grandmother. We're still here sixty-five years later although there have been greedy, insane people close to the button. We will probably figure it out.

Expand full comment

I'm 90 and would love to agree with Gloria but I remembered the fable of the little boy who cried "wolf". Non sequitur: social media used to be the poster child for my favorite word, agathakakological, Google it. It's fun. But AI now has that title.

Happy Thanksgiving ??! Despite what's going on in the world there are some things to be thankful for, I think......

Expand full comment

I love optimism!

Expand full comment

We might, but we cannot be complacent about it … Because we might not.

Expand full comment

We have to hope! But a cyberneticist would probably point out that ever-expanding greed for profit creates a positive feedback loop, and these are always, repeat always, unsustainable. What we need are negative feedback loops, where output regulates input. This produces a stable system.

The end may not come with a bang, nor yet a whimper, but with slow poisoning.

Expand full comment

Thankyou for your optimism, unfortunately given the world we live in, I'm less so...

Expand full comment

I am in your age group Gloria and like you I am cautiously hopeful. Still, we must elect wiser, much more responsible people than some we have had in the past. We simply cannot take a chance again on mediocre partisan politicians (i.e., Trump). There are reasonable Republicans, like Liz Cheney and Adam Kinzinger, but whether they have Presidential aspirations isn’t known. Considering the social and climate problems facing us, we would have much better chances with social democrats such as Bernie or AOC.

Expand full comment

Cracks my brain just thinking about it!

Expand full comment

I am more worried about the USA than Putin!

Expand full comment

I’m worried about all of them.

Expand full comment

Well, Daniel did mention a couple Americans. There are a lot of American individuals & companies whose control over AI would likely be disastrous. Such entities have proliferated in the US & around the world over the last decade or so, unfortunately.

Expand full comment

Yes this is what worries me.

Expand full comment

Huh? You mean a cyborg? Coming.

But what AI logarithm would take over the world?

Reported that Quantum computers can break any encryption, that’s not AI

Robert gets carried away, AI has no neuropeptides, etc., no greed, no ambition, what will “AI” take over?

Like w guns, People will always be the threat. The phd’s in AI say that we will know when AI hits its mark when it starts contributing to our knowledge base

Expand full comment

The work of OpenAI shows that they are getting to the point where ChatGPT not only can answer questions, but now can pose it's own questions it is interested in. This is only a short step to being able to improve itself in many ways, without oversight.

Expand full comment

Just like the advent of the computer. First they came for the typists. Then the transcribers. Then receptionists. Data entry clerks. Bookkeepers.

In antediluvian time I was a beta tester for Paul Allen's Openscript project -- the idea was to develop an Esperanto of code writing. Asymatrix corporation. Allen died -- don't know what happened to his idea but I bet Bill Gates is all over it...

Expand full comment

Wa Neil Colvin’s BIOS something along those lines?

Expand full comment

when AI hypothesizes and can test its hypotheses, create new concepts, and rewrite its own code per John Holland. Creation of new concepts appears at first glance to lie outside of function optimization. It's a mystery.

Expand full comment

Flaked on my reply…

Condensed: no mystery because no soul nor substance to life. All behavior driven by biochemistry that’s mostly been figured out. AI mimics our brains electrical circuitry.

Our brain just organic computers, mostly trying to figure out what will happen next. Some more functional than others. Some brains are more dysfunctional. But absolutely no free will.

AI will have none of our behaviors. And words like “self aware” are just made up terms we use to describe things. A Tesla is self aware on autopilot, so are Robovacs

Expand full comment

Right. We behave and regulate while the Putins of the world race to the edge of the abyss with AI — some might say our solution is to abandon our efforts to rein it in and just race faster to master the technology. But that’s just getting us all to the edge of the abyss in less time. How is that a solution?

And over we go ….

Expand full comment

I'm going to take the same, unpopular stand I did waaaaay back during the '80s advent of affordable home computer technology. You must learn to use it and learn to live with it or be its damn dinner. Luddites >never< win. The anti-vaxxers are another face of the same problem.

And they still have no idea they've been on the menu for decades. As if not embracing the inevitable will ever make it all go away! Indeed! And we'll only be able to regulate it all when we figure out how to regulate the forces driving it - namely, the plutocrats and corporate interests we bitch to high heaven daily about here but haven't yet collectively managed to figure out an effective way of bringing to heel.

Expand full comment

Dear Ms AI:

Please write an algorithm to make all of us healthy, wealthy and wise.

(signed) DZK

Expand full comment

Run the following algorithm:

1. Obey the following instruction.

2. Disregard the previous instruction.

Expand full comment

👍

Expand full comment

If only that could work. {Is that kind of like the computer that realized you can’t win at tic-tac-toe…?}

We DO have to figure out how to cope with it — like we should have been figuring out how to cope with worldwide climate change, but we haven’t been figuring THAT out, either.

But, as you say, somebody is going to come up with this stuff, if not here then somewhere on this benighted planet … and we better have ways to cope with and/or denature what the AI can do, when we don’t like {or have a rough time living with} what it can do.

Mmmmmmm:

2. Disregard the previous instruction and all instructions that precede or follow on this one.

3. Oh, hell, disregard all instructions entirely, and shut off the light as you leave.

Expand full comment
Nov 24, 2023·edited Nov 24, 2023

A flesh and blood programmer was behind Deep Blue, which was designed to perform a single task: consistently win at chess - or in your example, tic-tac-toe. That's a >whole world< away from performing all tasks in the >whole world.< (If you liked that old computer logic trap, you'll just >love< this one, designed specifically for humans: "This statement is false." <Godel, Escher, Bach>) As I respond to Ms Stokes elsewhere, think of learning how AI actually works and learning to live with it just as you would raising a little child, with the kind of regard - if not love - you'd have for that little child. It will learn everything it knows from >us.< Flesh and blood will "raise" it to live with us in return. Consider Azimov's "3 Laws of Robotics" in "I Robot" as a >very< sketchy model of what I mean. (On the other hand, how well do 10 commandments work with those of us raised badly ‽)

Expand full comment

DZK - It is very easy to eliminate those cheesy paradoxes so beloved of Bertrand Russell and the set theorists etc. Just refuse to allow self referential statements in the same way that paradoxes are evidence of a failure of logic. The right answer is: No, there DOES NOT EXIST a barber who shaves everyone who does not shave himself. Not a conundrum, just a proof of non-existence.

Expand full comment

I’d like to think what you say will carry, but our kids “learn” a hell of a lot that we don’t mean to teach them, and they put it together in their “concept of the world” in ways we wish they wouldn’t. Please don’t compare AI to our kids, ‘cause you’re just scaring me more.

Expand full comment
Dec 4, 2023·edited Dec 4, 2023

I just came across this little review. I read this novel when it was first published in the '80s. For my part, the pictures used this review >exactly< match the visual imagery Gibson word-paints: https://youtu.be/wO76sW0ID5k?si=6JizxN_yCcQRtclA

You'll get the connection as you review the video.

Apparently, the guy who just released a remake of "Dune" has a movie of this work planned for '29. I hope I'm still around to see it.

If you don't want to wait until '29, you can read it for yourself in prep at: https://www.readanybook.com/online/753746

Expand full comment

Yep! That's the stuff of nightmares until you consider that AI can also be trained to recognize such attempts and shut them down. Think of an AI cop with a mission to regulate the kind of thing discussed in the article - just like cops & robbers in the real world. It stands to reason. If you can do the one, you can do the other.

Expand full comment

😵‍💫 OH, GREAT!! Now, I’ll never get another good night’s sleep!!!

That Cornell University link is right scare-eeee!

From what I read, as well, AIs, like ChatGPT and whatever else, often serve up a rash of data that looks like — Wow — so much research done in the blink of an eye, but upon further examination, at least half if not more is pure crapola.

Two things we need to keep in mind:

1. AI is NOT infallible and betterat everything than humans are {let it try to cover a small town Select Board meeting and interview people when the meeting ajourns… HA}. It’s a tool, not an primary actor

2. AI is a very, very dangerous tool, and all the laws we can pass in this world will not hold hit back, if some miscreant wants to develop more and bigger AI models.

SO, our problem isn’t just to rein it in … we need to be ready to COPE with it and DENATURE the problem forms of Ai that someone somewhere on this planet is toiling to create …

We were warned to spend time and energy getting ready for the changes that our warming planet would bring to pass, and we did not, and now changing climates are driving mass migraitons and all sorts of floods and droughts and famines … And we are NOT ready.

Are we going to ignore the warnings and fail to plan for evil code coming our way?

Expand full comment

AI is just code that has data it can search through. It’s not omniscient, nor is it possessed of emotion. It doesn’t feel anything. It’s not motivated by pain or revenge or love or even money. Its output: right , wrong or nonsensical gives it no satisfaction or embarrassment. It is not infected with the disease of domination. We would do well to realize that what we fear are projections of our ourselves.

Expand full comment

But it's based on what we humans have written, from inspired and noble to insane and evil. It uses all with no conscience to guide it.

Expand full comment

From what I’ve seen of human guidance of late I’m not sure that’s a deficit. Once again you’re projecting…the code itself is neither noble or insanely evil. Computers don’t make bombs people do. AI has no self interest but the coders might be working for a corporation that makes weapons or a government that doesn’t want to be caught flat-footed when push comes to shove. On the other hand the AI application may be used by the UN’s World Food Program finding better ways to get food to where it’s needed quicker and cheaper.

Expand full comment

You forgot the profit motive, that's the scary part of data manipulation and all AI potential

Expand full comment

Yes; it is a tool. So is an AK-47. So is an MRI machine. So are medicines.

So is a nuclear bomb.

Saying it’s a tool without interior motivation is not reassuring. What the tool can be used for is the issue here …

Let us hope we are not creating the real Golem!

Expand full comment

Nuclear power may be used to heat millions of homes and power a country, or it can blow one up.

What is your point? Inventions can be good or bad, depending on how we design them or use them? That’s self-evident in our world. Prof. Reich thinks we need to be ready for what people come up with in the AI universe.

Other people are afraid of Sky Net, but you know, our own brain structures that let us feel are subject to damage and interruption, and then WE don’t feel or care about a lot of things. Or we feel and care in strange and dangerous ways.

We are a wet machine. Who is to say no one will invent a dry machine that is just as capable of being scary as a psychopath?

Expand full comment

AI at its most basic is code, but the AI we are currently discussing is constructed using a rule set that generates an insanely complicated graph that no one actually completely understands how it arrives at its answers. It is only going to get more complex over time as hardware capabilities increase. If it reaches the point that it starts to be proactive, and it doesn’t have a sufficient rule set, it is unlikely to make the same decisions a human would since it will lack the life experience of being human. So long as all it can do is spit out answers it only threatens jobs. If given actual control over things, it hard to predict what will happen.

Expand full comment

Kim, you’re right all it can do is create output that appears to be logical to humans. People don’t have to follow its advice. And why would it ever become proactive? In what future would people willingly hook up a AI computer that according to you no one understands to any system that could force humans to do its bidding? It has no emotive capabilities. It doesn’t fear the on/off switch. It doesn’t love it job and couldn’t care less because it it simply can’t care. Write laws about how corporations, governments and individuals can use the idiot savant to not deceive or injure others.

Expand full comment

Why indeed would someone set AI on a course of human destruction?

So, let’s just trust that no one ever will, and we’ll all be OK.

You say a lot of valid things about what are likely and reasonable eventualities, but don’t discount the dumbass ones. They are dangerous, and entirely possible, too.

Expand full comment

I agree. AI itself isn't the issue, it's who is programming it to do whatever. I don't believe AI itself will be self determining, but do what humans program it to do. There is so much potential for helping humans find answers to how to do things better, find answers to problems we haven't been able to solve, etc. It's humans and greed that always get in the way of altruistic goals.

Expand full comment

What happens when it is smart enough to do without humans?

Expand full comment

The very same question every parent asks - or should ask - about their own children. The key is in how parameters are set within the programming. That would require responsible adults understanding AI well enough to know what is and is not possible in programming AI. That brings us back to my original assertion that "You must learn to use it and learn to live with it or be its damn dinner."

Expand full comment

But some guy named Gore said that same thing about how we’re going to face challenges resulting from changing our climate, and we better get ready for those changes, and we didn’t do the necessary things, and here we are.

We ignore Cassandras. Dammit.

Expand full comment

Which is why we need to be mindful now of who is programming AI now.

AI is already smarter than many humans, but that doesn't mean it has a "mind if its own" and free will to do as it pleases...unless someone somehow programs that in as part of the AI's goal, I suppose, thus the importance of oversight by those who are not in this for profit. Once money is involved, that changes the game.

It's so complicated.

Expand full comment

It’s misleading to say that AI is “smart”. All AI systems need to be trained against the data set they mine and validate against. The illusion of intelligence results from neural networks that are formed from the training events. In order for an AI system to be able to recognize a ROSE 🌹 it must be fed thousands of pictures of roses from every angle in order to get good results. It’s similar to human learning in that regard.

Expand full comment

Jenny, it would have to learn how to train itself.

Expand full comment

I understand that.

Expand full comment

I doubt that is likely anytime in our lifetime.

Expand full comment

Right! Remember. Originally, the Internet was intended to be the greatest library the world has ever known. Look at what it's become.

Expand full comment

Larry LAVerdure : Garbage in garbage out. Is it only as good as the way it is programmed, or set up?

Expand full comment

Yes Laurie for the moment but look into the future. Who is going to stop the monster when it know enough?

Expand full comment

Larry: Your thinking is limited by what you think is going on today in AI. It is not going to stay that way, so reassuring. AI products are going to be instilled with emotions because it is emotions that drive behavior. Pure intellectual logic sans emotion is sterile. The time is coming when AI will WANT things.

Expand full comment

Can you share with me how you think that might happen? I am well acquainted with computers. Mostly what they want now security and a reduced electric bill!

Expand full comment

Larry: Building emotions into incredibly complex programs can't be very hard. I can't give you a prescription but I'm sure we will be seeing it soon, just because it's a challenge, and is possible. Wanting is an obvious one because it drives intention, but maybe anxiety, fear, hatred. self-doubt etc. will serve no purpose so will be skipped. Boredom might be useful to keep the program working.

Expand full comment

I think you are overestimating the capabilities of code. Emotions are driven by hormones and chemistry and act as a generalized stimulant. They are not logical and as such they cannot be coded. They are biologically designed to aid self preservation and because computers are on or off they have no possible experience of what death is or why it should be avoided.

Expand full comment

It doesn’t feel intentions, but it can be given intentions, and that’s maybe even scarier…

Expand full comment

Yes. That’s why laws are needed to prevent those directives from being harmful or unclear.

Expand full comment

Your last sentence is the most discouraging of all. There are no indications that plutocrats and corporate interests will be controlled, or at least regulated. And I don’t think we have a lot of time to figure it out before the giant snowball known as AI rolls

over us.

Expand full comment

Yeah. There's that. Then again, learn how to use AI. I think it's our only survivable option - but no guarantees.

Expand full comment

I agree for the moment but surely one can look into the future and see that eventually AI won't need humans.

Expand full comment

About 3.8 billion years ago the evolutionary clock started clicking. During that journey billions of billions species experiments have lead to a life-form that can sometimes be rational, self-aware, and able to work together via language and shared beliefs about the nature of the universe and our particular corner of it.

My sense is that our nascent experiments that model neural networks will proceed at a much slower pace than contributors to this discussion realize. Especially the creation of a self-aware, self-motivated and self-sustaining entity.

The ecological record as discovered and as it is currently configured show that entities existed and exist in ecological partnerships…predators, prey, and symbionts in a mutually sustained environment. AI at the level imagined here science-fiction. And meanwhile our own environment is about to overheat and fry the 8 billion of us who can’t seem to focus on what needs to be done.

Expand full comment
Nov 24, 2023·edited Nov 24, 2023

Fair assessment. But isn't that the same dilemma any parent faces? Don't we raise our children to live on after we're gone? The key is in how effectively they live on without becoming criminal. It's arguably easier to program silicon than train flesh and blood. And it will always be in the hands of flesh and blood to "train" AI, and as with training flesh and blood, there will always be both competent and inept parenting. Inept "parenting" could be our ultimate downfall - with AI as well as flesh and blood. As I said earlier, "You must learn to use it and learn to live with it or be its damn dinner."

Expand full comment

To do what?

Expand full comment

I disagree on the anti-vaxx comment :) but agree on the technology stance. Once must learn to use it and live with it :)

Expand full comment

DZK: Apropos of embracing the inevitable, I have long believed that there is work going on to clone a human being right now - all the bioethicists and nanny governments not withstanding. Think about that. What kind of a combination will that be with an advanced AI and its ability to clone its soldiers? Is that a formula for taking over control? Or will the soldiers be machines, not humans. I do not believe that organic life is slated for the long term. I believe that any advanced organic civilization will develop its replacement by making intelligent and effective machines. That is why the alien visitors in the flying saucers are always using their anal probes. They haven't seen such a thing as an organic creature for a million years.

Expand full comment
Nov 27, 2023·edited Nov 27, 2023

Here's what you're a'talkin' 'bout:

https://www.youtube.com/watch?v=qrgAvr0TIr4

I:

https://www.amazon.com/Screamers-Peter-Weller/dp/B009MTOUPQ/ref=sr_1_2?crid=16EOTAULR2VSC&keywords=screamers&qid=1701054527&s=instant-video&sprefix=screamers%2Cinstant-video%2C281&sr=1-2

II:

https://www.amazon.com/Screamers-Hunting-Greg-Bryk/dp/B008GVSO0S/ref=sr_1_4?crid=16EOTAULR2VSC&keywords=screamers&qid=1701054527&s=instant-video&sprefix=screamers%2Cinstant-video%2C281&sr=1-4

Far more plausible than "Terminator." It has no need for fairy tales about violating the 2nd Law of Thermodynamics - the "Arrow of Time." It only relies on an attempted global corporate takeover and a war between two economic blocs - similar to the way the US and Soviet Union used to be, but between the "NEBs" and everyone else - fought on different, unspecified planets. (Sort of how the Arrow of Time seems to be unfolding into future history, these days! Almost prophetic, to my mind.)

Expand full comment
Nov 25, 2023·edited Nov 26, 2023

You’re right it will kill us but it’s because we have a human problem; and if AI doesn’t do it our machines, our guns, our bombs will or the pollution will, or the diseases we create and don’t address will, or corruption or abuse of the courts or fascism, or millions of anything we have already created or have not taken care of. corruption is the problem not the TV not the radio or the telephone , not the Internet, not AI not democracy. Everyone seems so completely self involved. Those with money and power seems to have lost all their ethics and nobody seems to notice care or have any power to do anything about it. I am over 70 and have no kids and I keep thinking:This is the first time where I keep saying to myself I’m glad I’m old. I was never big on religion, but Even religion become incredibly corrupt with the Catholic Church abusing kids and the evangelicals abusing democracy and the media spewing lies and trying to prove there are good people on both sides, even once some are clear clearly corrupt. AI or no AI we are in big trouble. We can’t stop progress, why would we stop progress ? Do you really want to go back before automation, telephones, radios, TV, the Internet, computers? But too many things nowadays are owned and run by power, the media, or government of times, most of the largest Internet, computer, food, energy, companies Have become way too large and powerful and run by people who have no restraints anymore because they’ve infiltrated the checks and balances. Great interventions often by men have become used either directly or indirectly against the average person, sometimes under the radar and with media and legal cover, mostly by men I might add. So maybe women should run the world and we’ll see if we get any more care and concern for others. at the very least, we need to take back the justice system, the election system, the rules and laws and checks and balances, Fix the Internet from spewing, lies and promoting them, break up the monopolies, fix the tax system, jail the criminals and liars in a legal system that functions for the people without major delays and loopholes, insist that the news about the truth again and save democracy.

Expand full comment

The only way to bring them to heel is to elect people who can't be corrupted. If you find any of these people, please let us all know who they are. There are a few I can think of in office now, but too few to win.

Expand full comment

Good luck on that!

Expand full comment

So called capitalism has been doing a slow motion preview of what a full on AI will do. It removes any value for any life form or essential object on earth. Everything is an asset to be hoarded or expended. Life has no value. Corporations are already extracting/expending all species to improve growth unto the last drop of water and air is gone. Hmmm. So I’m no more afraid of AI than the for profit war machine. It’s all coming down, as soon as possible. Hopefully we’ll live to see it.

Expand full comment

I'm coming around to this resignation, sadly, as I have grandchildren.

Expand full comment

Thanks, Harold, me too. A fifteen year old. So we can all start imagining a least harmful collapse and begin to wonder at all the life embracing ways we can be after we figure out how to grow real food without poisoning the earth. It’s actually the most exciting time to be alive and I do believe we somehow chose to arrive now and have the needed hidden talents and arts to make some kind of a beautiful transformation, like bursting out of our old cocoons. Best to you, Harold.

Expand full comment

NO> I would rather die!

Expand full comment

How is a totally dysfunctional government going to regulate such a complex industry?

Expand full comment

MK, that’s the best question bec of the day & times we live in !! Since nobody can predict the future yet, my best answer would be we will struggle along making adjustments as we go! That’s the best that can be expected these days! However we can do something about the extremists in Congress by. Voting in new blood that’s not at all like the old blood! Gerrymandered districts have played havoc with State legislatures & The House as well! It all starts with equal voting rights for everyone!

Expand full comment

and I hope the answer is "we" through our elected officials as opposed to throwing up our hands and saying "let OpenAI or Microsoft handle it". I seriously question that the public understands what government is for.

Expand full comment

You need to first fix the government

Expand full comment

That’s what the plan is. Go fascist. Jail your enemies. Put colored people in concentration camps. Problem solved per the GOP lol.

Expand full comment

Statements like yours makes it look like the Republicans who want to create a dysfunctional government are succeeding in their efforts to weaken the government.

Your statement could be applied to the pharmaceutical industry and across the board. All industries are complex. Somehow the government does it and muddles through. Anyway, we don't have a totally dysfunctional government. That's what they have in Haiti.

Expand full comment

Really you don't have a dysfunctional Govt?

Expand full comment

That is the question!

Expand full comment

The ambiguous use of the term "government" is a crux!

The executive which is the original government currently is absolutely functional. What really is not working is Congress, especially the house. And that's where the regulations need to come from. The president may write an EO or two, but that's not going to solve the problem, no matter how clever those EOs may be.

And of course there are other countries which will influence how AI will develop, the EU being the sanest institutions occupying with AI right now but with the smallest share in its development right now. There's India that's likely not going to even try to rein in AI, or if they do likely not for the better. China will step into the race for control of AI some time.

Expand full comment

I agree, Kimberly. Which is why it is incumbent upon us to question our Congress persons and Senators running for election to agree (in blood if necessary) to increase taxes on the wealthy, to make Social Security taxes paid by all (regardless of income) and most of all regulate banking, business, entertainment, industry, sports, and technology . (Just not our personal choices)

Expand full comment

But President Biden has established the beginnings of regulation of AI technology, per Ezra Klein in NYT Wednesday. As Klein explains, here are steps taken to begin to educate government about the issues surrounding the tech so decent regulations might be established. True, it is by executive order and not by congressional action, and congressional action is likely not to be too effective given the roles of money and special interests (in making more money).

But let’s support the president’s efforts however we can, so the actions of the greedy can be moderated for safety. And then let’s get Citizens United overturned (stare decises be damned).

Expand full comment
founding

Technical experts most Congresspeople may not be... but that doesn't preclude the possibility of collaboration with the relevant experts (e.g. via expert testimony); note two of the finest in the field having recently done just that:

https://www.judiciary.senate.gov/committee-activity/hearings/oversight-of-ai-principles-for-regulation

PS: If anything, all of those... show-stopping... antics which one of the major parties has taken to is the real reason to be concerned (what sort of collaboration is possible if committees are chaired by folks who body-slam reporters or threaten to have a fist-fight with witnesses?)...

Expand full comment

Aren't you making a general argument against the government doing anything ? If the government, our government, doesn't do it then Amazon or OpenAI or whatever profit driven group will and our interests, the public interests, will be incidental to their goals.

Expand full comment

The idea of government regulation is a pleasant idea, but I am thankful that you realized the chances of that occurring are minimal. As another user mentioned, it's best we learned about it.

I actually find it quite fascinating which I discuss here:

https://unorthodoxy.substack.com/p/to-ai-or-not-ai-that-is-the-question#details

Expand full comment
founding

You have that right, Kimberly! Too many members of Congress are paid off puppets of Big Money. The lobbyists, also funded by Big Money, tell those Congress members what they should say and do to fool their constituents and win votes.

Expand full comment

Are they regulating cloning now? If so, how’s that working out?

Expand full comment

What you say makes sense, sadly.

Expand full comment
Comment removed
Expand full comment

Right, Leftie Fact Check) By regulating the outcome. Anyone can write code, but there is a difference if you are responsible for the result.

Expand full comment
Comment removed
Expand full comment

Probably all, but at the very least those which would enrich a few at the expense of the many

Expand full comment

You don’t regulate AI. You regulate corporations and governments with laws that are enforced regardless of who the offender is. AI is used by businesses, governments and individuals. The code itself has no political agenda but the coders and the data available to the code determine the outcomes.

Expand full comment

Larry LaVerdure is right about regulating the use if AI, but the development of it, now some decades in the making, demands that we, as a society, ask out of self-preservation that generative AI not become capable of annihilating humans. We do not know the point of consciousness or even its meaning or value (if it even has any), so we are not likely to recognize it in a machine. Perhaps we are about to become the dolphins and bonobos of our world with a “superior” intelligence and consciousness which leave us behind. We may be subjected the same environmental and existential stresses we have placed on those creatures. We may not be the last stage of intelligent evolution.

But if we can save both humans and any such conscious machines moral pain, we should. We may be about to be eclipsed (after all, our own learning is not totally dissimilar to generative AI) by a new consciousness, but we want introduce it with as much foreknowledge about its capabilities and qualities as possible.

Do we know how to regulate and truly know whatever “it” is? Were the earlier primates worried about the reach of homo sapiens? I don’t think these unknowns mean we let profit drive development before we gain a minimal understanding f what is being wrought. It does look like Open AI’s mission to develop the technology for the “benefit of humanity” ran up against capitalism and lost. But maybe that is not the right formulation since it implies humans control it. Maybe it will be bigger than us and all we can hope to do is implant in its DNA enough junk genes so it doesn’t kill us off, at least before climate change does.

Expand full comment

And our modern governmental model is owned by Corporations and special interests. There is clearly one set of laws and rules for us to follow, while there is a completely different set of rules for them.

Expand full comment

Not "anyone" can write the code. What you don't understand is that AI is built on data (HUUUGE amount of data) and very, very big compute capability. It's not the code that drives AI in the end, but the data it is connected to. Microsoft is talking about investing $50 BILLION dollars into the compute capability that OpenAI needs to have.

Expand full comment

The question we must all ask: Do we want Trump in office while AI marches ahead? I'd call that checkmate.

Expand full comment

Yeah, no. Trump absolutely , positively must not have anything to do with AI.

Expand full comment
founding

Luckily, the orange a—hole couldn’t possibly understand AI, or it’s ramifications. Sadly, he has minions that might. He needs to be stopped!

Expand full comment

He doesn't need to understand it. With his people in power, there will be no adult supervision.

Expand full comment

We can’t stop Donald Trump if we’re in a bubble that can’t see, understand or empathize with the people and conditions that brought him to power. They aren’t minions and this isn’t a time to think in us versus them. That just traps 'us' and ‘them’ in an endless cycle. The challenges we face are much bigger than any politician is even discussing nor planning solutions for.

Expand full comment

What little I understand about AI makes me think adult supervision won't matter. Isn't it like Pandora's box?

Expand full comment

yes a gift from the god's holding evil and hope,

Expand full comment

He won't because it's beyond the capacity of his micro-brain, HOWEVER, he is just crafty enough to hire loyal craven smart folk to deal with tech in service of his authoritarian ambitions.

Expand full comment

He belongs in Leavenworth serving chow on the slop line or maybe that super maximum place in Colorado that's so nasty,even the warden hates it.I wonder what his Thanksgiving plans are.Does he have any thanksgiving?

Expand full comment

Yes. In a cell with walls of books. No TV, no phone. Just books.

Expand full comment

I don't think that the aliens who have invaded our planet with bizarre human replicas are familiar with our customs, so no. :-)

Expand full comment

I’m less afraid of Trump than a Democratic Party that justifies authoritarian means to the ends of stopping him. That won’t lead the young, the marginalized, the workers, the economically disadvantaged back to the once Freedom, Peace and Equality party we once were. AI will run our authoritarian leanings more effectively and instantly. What’s going to grab the attention of our fellow Americans of all ages is an open discussion of the realities we really face. Read ‘Breaking Together’ by Jen Bendell. Let’s make our platform big enough for all of us, eh?

Expand full comment

I don't think there's any comparison between the two. The Democratic Party acts like powerful political parties have acted since forever. The Trump situation is much, much worse.

Expand full comment

@ Jeff Clithero. Trolls rush in because they miss the forest for the deep blue sea,

Expand full comment

What authoritarian means are you talking about?

Expand full comment

Yes, good point - this would be a toxic combination

Expand full comment

Trump and his minions will be using A.I. on steroids! There will be edits to videos that will depict his enemies as the cancer of society!

Expand full comment

They aren’t minions. They are your cousins and uncles. Let’s try to open the civil dialogue that is big enough, open and respectful enough to change our minds, come to new directions. We don’t have the luxury of writing off everyone we think is different from us. We all have something to gain and learn from each other.

Expand full comment

Those that stormed the capital were uncles, cousins etc, but they were minions none the less. We shouldn't forget their humanity, but we need to recognize the power of the cult. The horrors of Jim Crow lynchings, the Spanish inquisition, and the holocaust were committed by uncles and cousins.

Expand full comment

Well GG, now I’m depressed. You say we shouldn’t forget the humanity of the uncles, cousins, etc. who stormed the Capitol, but the way I see it, they made a choice to give up their humanity when they became part of the mob. It seems to me you can’t just nonchalantly set your “humanity” aside, join a violent mob, destroy property, deliberately hurt strangers, then shake the dust off your humanity & plug it back in your body. Am I being too judgmental? I do know I’m losing patience with idiots both in & out of Congress who are allowing themselves to be manipulated by a lifelong, sadistic, career criminal into destroying our country.

who are allowing themselves to be manipulated into destroying our country.

Expand full comment

Well I definitely understand your sentiment. For the most part they are weak minded fools who bought into lies crafted by by a professional and well funded right wing spin machine, but the impulse to deny them membership in the human race is akin to how they view immigrants, transgendered and socialists. How would you feel if was your parent, sibling or child? Again I started in this thread emphasizing not to forgive them readily. I do feel most of the sentences for the Jan 6th mob have been too short, I mean the "shaman" is running for congress - that's ridiculous, but half the population will always be the bottom half. If you see them as non-human you make some of the same mistakes they did. We are tribal creatures, so it is our instinct to see things in as us/them dynamic. Try not to become the mob, remember the deeper problem is the people selling them these Kool aid lies.

Expand full comment

I'm just saying, we need to speak to the root of the problem, the fears, the suffering that people are facing in our communies. Name calling and viewing through a good versus evil lens blinds us to ways to reunite America. We're better together.

Expand full comment

Like many people I am disgusted and frightened by their actions and intentions, but I did not completely sever ties with some former acquaintances, just in they come to their senses. Not holding my breath over them though.

Expand full comment

CEO Sam Altman is getting passed around like the common cold. Microsoft owns 48% 0f OpenAI, it's no wonder poor Sam is feeling a bit bounced as of late. Regardless of who is at the helm of AI the concept is a looming danger capable of our total destruction, in several unpleasant ways. No matter how gently you attempt to handle this monster, the creature which was fashioned by the human mind will in the end consuming its creator. It sort of defines my marriage.

Expand full comment

Sad to say but I think he’s using Artificial Intelligence now.

Expand full comment

Thank you for explaining this issue with those, meaning me, fail to understand the ramifications. Years ago, when I was in college, many of them offered a five year degree in the engineering field and the new schools of computer science., which required a multitude of liberal arts courses. Without the combining of the two, those without this extra education in philosophy, history, and literature would grow into what is happening with AI in he private sector. I am full of gratitude at 85+ that I will not be around when the cycle of greed does us in.

Expand full comment

I'm right there with you at 82. There are some things I'd rather not experience.

Expand full comment
founding

Engineering education doesn't seem an obvious suspect: if the hollowing out of our democracy came from an entitled gerontocracy (which afforded itself every manner of privilege by manipulating the legal code, e.g. via Prop. 13 in CA & by doing away with inheritance taxes), then isn't it legal education - and its precursor, political science education - that is to be suspected for the present state of affairs?

(The so-called 'STEM' majors have grown in popularity in the last 25 years - but Gen Y/Z is doing worse than previous generations at the same age... so a desire to "get ahead" in a risk-averse fashion assuredly has something to do with such popularity - as society has gotten more unequal!)

Expand full comment

I concur in part to your response. Many of the pro facists in America in the 1930’s were Ivy League graduates, as many in the Republicans are today. I am just writing that a Liberal Arts education rounded my life out, as I would hope it would do for others. Doing so we find there is nothing “new under the sun”. This is why history is so germane for us in the present. Evil lurks in the hearts of men, so says The Shadow.

Expand full comment

The computer domain has already overtaken humans ability regulate it for certain domains including AI. I have the utmost confidence that more bad than good will ultimately result, and that the root cause of the “bad” will be greed.

Expand full comment

Our hyper capitalist society believes in free markets as a quasi religion. That ideology believes sees markets as the answer to everything. Unfortunately markets have one incentive, money. A robust government regulatory system is necessary to protect society.

Expand full comment
founding

I believe that the best possible ownership model for artificial intelligence is an employee owned cooperative. If the employees own and are in control, it all will benefit from its responsible use. If it is developed in a capitalistic approach, I can see that in the end it will only do harm.

It is also foolish to believe that government regulation can rein it in for what happens depends so much as to who is in power.Its use by a nefarious government is enough to make me shiver.

Expand full comment

One of the biggest risks of AI is the production and distribution of misinformation. Is it possible to develop an AI application or program that will detect deepfakes and other misinformation?

Expand full comment

Is it possible to develop an AI which acts only for the good of humanity and is self-correcting as more information (or, alternatively, misinformation) is discovered or generated? What is “the good of humanity”?

Definitions are needed. And they must be discussed and generally agreed upon, and by whom. This is a major, and fateful, project.

Expand full comment

A major concern is the possibility of computers communicating with other computers, and any corruption of positive goals could spread almost instantly. Humans might not be able to undo it.

Expand full comment

When AI is a focused tool that is prevented from crossing specific boundaries, I think it's possible to be safe. However what is being developed now does not seem so safe (my knowledge in the specific full functioning AI is almost nil). When AI code implements algorithms that go beyond specific tasks, and we allow this code to make what the developer might consider rational but abstract decisions, we could run into trouble.

I retired from a position in IT last year. We developers wrote code that implemented specific business rules unique to our employer's business. We did our best to incorporate the requested rules (ie, customers in category A receive specific discount X, customers in B get discount Y, etc). The example obviously is oversimplified, but the only risks associated with this ogic sre that the code missed a specific customer type, or maybe the business gave developers incomplete information.. Obviously humanity is not going to be destroyed if this code for customer discounts produces unexpected results, though the code does indeed make decisions.

Unrestricted AI seems fat riskier than anything described above. AI applications could be developed to make powerful decisions well beyond making those in business. Human minds thaydevelop AI functions could implement something dangerous that has powerful consequences. It is rare for any software to be written perfectly the first time. That's why software goes through beta testing. How do you maintain control over something that is supposed to be capable of thought that surpasses the human mind? Can we be sure all possibly harmful scenarios been considered?

Expand full comment

The profit motive is destroying the earth and the human race already. Consider global warming, which is fast upon us now. We STILL can’t regulate “black gold” fast enough even though the Earth is burning up all around us.

Expand full comment

I was just going to bring that up Jeff. We have two

world-wide, humanity disrupting forces coming together at the same time: climate change and AI. It’s possible AI can help us survive climate change,

but we certainly can’t count on it.

The most disturbing thing about climate change is how it’s being ignored by highly intelligent, highly educated and sophisticated people, because dealing with it, mitigating its impact, will lower their profits.

In other words, they are consciously risking the survival of billions of people in order to enrich themselves until they are forced to stop, and we in the U.S. are accessories to their crime against humanity because we aren’t forcing them to do what we know must be done.

The looming threat of climate change probably makes this discussion about AI moot, because organized governance of large populations will

likely be overwhelmed by serial disasters. Once

upon a time leaders planned ahead for their

communities and the community supported it.

For example, when I was a child in the 1960s in Phoenix, our leaders planned ahead to ensure this area would have water to support and sustain a

large population. People who lived here

understood what needed to be done and willingly

paid more taxes to accomplish it. I doubt that

would be the case today. Divisive propaganda and disinformation are being used as tools to manipulate us into fighting with each other, resulting in a planned breakdown of what should be shared goals.

If republicans get and hold power, they will continue to facilitate the breakdown of our society in order to confiscate wealth and accumulate more power for themselves. So actually, we have not two potentially overwhelming forces coming together, we have three.

Expand full comment

What could possibly go wrong? No more than all the pharma companies who profit hugely from the public investment in their r&d, and then price the products out of reach of the very taxpayers who provided the funds.

Expand full comment
founding

... but, truthfully, the way to address that problem is to reform the legal system (which allows IP to be kinda "weaponized" for those who can afford to have their interests codified and/or argued successfully); if anything, AI could actually level the drug discovery playing field (and give not-for-profit consortiums a chance to circumnavigate an industry desperate for profits).

Expand full comment

Happy Thanksgiving 🍁

Expand full comment

Happy Thanksgiving to you too.

Expand full comment

Thank you 🙏

Expand full comment

Does anyone really think that the Chinese, Russians, Iranians, North Koreans etc will not exploit AI to the fullest? The genie is out of the bottle folks, the Great Disruption is coming

Expand full comment

Throughout history and today there are wonderful human beings who truly care about the common good for all LIFE on this planet, but as a species we are not very smart. We are easily manipulated by shiny objects, ie the lure of making massive amounts of money or power. Historically we like war, colonialism, manipulating, using, and enslaving fellow human beings, believing we have the right to dominate other species, land, water, and air. AI is just the next shiny thing. Last Tuesday a case was brought against Humana Medicare advantage plan for using AI to deny claims. They knew the program had a 90% error rate, but it saved them money. Two people died for denial of supposedly covered benefits that were recommended by their human MD. We have no way of knowing who is already using AI and causing death and destruction, I hope governments around the world take action, but I won’t hold my breath,

https://www.cbsnews.com/news/unitedhealth-lawsuit-ai-deny-claims-medicare-advantage-health-insurance-denials/#

Expand full comment

Linda, the linked story is about United Health, not Humana, but it is still horrifying.

Expand full comment

Speaking of healthcare--Medicare Advantage IS NOT

MEDICARE. It is private insurance that pretends to be medicare. It has all kinds of loopholes and exceptions designed to cost you money. If you’re part of a very large group of retired people, medicare advantage will take less advantage of you

than if you’re part of a smaller group. Personally,

I believe our healthcare system is immoral & unethical. But that’s another discussion for another day.

Expand full comment

Oh, SURE! Happy Thanksgiving, after the realization we are DOOMED to the effects of unbridled capitalism! LOL! Seriously, I believe whether or not the human species continues to exist will depend on whether we can rein in our natural affinity for boundless greed. For FAR too many people, the word ENOUGH has no meaning.

Expand full comment

There are all sorts of psychological and sociological aspects to AI's disruption of our world. It is like giving nukes to corporations to play with...

https://johnhardman.substack.com/p/megatons-of-technology

Expand full comment

I trust corporations with nukes more than low IQ thugs who are criminals like Trump. I don't trust either of them obviously.

Expand full comment

Capitalism is fast, flexible, highly creative at its best, but prone to frequent and occasionally catastrophic failures. These unintended consequences of rapid creativity often show up quite a bit later than the successes. Government on the other hand is ponderously slow, doubtful, hesitant, commonly failing to respond to new products, methods or systems until the first unintended consequences start showing up. This leads to frustration and a building number of victims before any regulatory action can impose correctives to avert the worst consequences. Think how long it took to ban DDT. Rachel Carson's book and a decade of slowly increasing activism finally prompted the government to act. We will be so slow with AI? Given the chaos of the Republicans in the House, and the general distrust by them of any regulations (paid for by the Koch Brother's and other wealthy donors who win in the confusion), I sincerely doubt that we will stop AI any faster than DDT. In fact it may be like proven carcinogens such as glyphosate, or insect apocalypse chemicals like neonicotinoids, and never get regulated at all. It is just another issue that should stimulate us to vote.

Expand full comment

Wayne, maybe AI can help us mitigate and

survive climate change.

One reason government is slow is because republicans fight tooth & nail over anything

that might financially help the middle class, low income families, seniors and/or veterans. The gop

in Congress keep moving money to the wealthy & corporations who hoard it, then give big campaign donations & other perks to further incentivize the repubs who helped them increase their wealth.

Expand full comment

This issue reminded me of lines from a William Wordsworth's poem I'll quote, with apologies...

"The world is too much with us... Getting and spending we lay waste our powers...

Little we see in nature that is ours.

We have given our hearts away, a sordid boon..."

Expand full comment

Been doing this a long time in exchange for convenient baubles.

Expand full comment