You have it spot-on as always, Gary. If you find it as unlikely as I do that our current government can manage any of your three criteria, you might be interested to know that the newsletter/podcast/movement I'll be launching on June 29 will provide a way for the people to demand a government that can make good decisions for us, if enough of us speak with one voice.
I'll be launching a newsletter/podcast/political movement called Unfucking The World, with the explicit goal to use a novel legislative strategy to force an amendment through to our Constitution in less time than anyone imagined possible.
The amendment is drafted, the strategy is set, (most of) the podcast episodes are recorded, and my savings are sufficient to see the project through to any successful conclusion: it'll either work within a year, or it will have lost momentum and failed. (I'm not accepting any money from anyone, to avoid accusations of scam.)
UTW isn't going to address AI policy on its own: it will give us a government that we can trust to make decisions about AI policy for us, and one that (I imagine) will be much more likely to take your concerns seriously.
The current situation suggests that we need to name a REAL distinction that gets blurred when we are discussing ideas about "the government."
There is the drift in thinking of "the government" as (a) a singular ideology aimed at having more and more power, and leaning towards unmitigated USE of and control of The People," e.g., the unitary executive authority idea fostered by the Heritage Foundation and its control-freaks; . . .
. . . and (b) "the government" (referring to a democracy in the case of the United States) thought of as (recovered as meant originally) a dynamic, working, civilized, intelligent, and reasonable set of true representatives of The People openly working through these problems for the good of all concerned.
The first interpretation of "the government" we are and should be afraid of, including (and unquestionably) a tendency towards PRIVATE economics holding all the power cards; the second we need to understand ourselves and everyone else as a working part of what needs to remain DYNAMIC and interactive, and more on the side of ideas of the dynamism of what needs to remain in the PUBLIC DOMAIN, including of the economics that underpin everyone's well-being.
When one person can blow up the world, we know we have done something wrong.
Every social system (whether a family, a company, or a state) is based on—or develops and changes based on—the information, communication, and interaction processes of the individuals involved. The purpose of digital information and communication (I&C) and AI systems is precisely to automate, monitor, and control these processes, thereby steering a social system in one direction or another.
One direction could be a democratic system; the other could be an autocratic system or a dictatorship. The fundamental question this raises is: How must I&C and AI systems be designed to steer a social system in one direction or another?
I would phrase the question differently: What rules or self-regulatory mechanisms do natural systems, for example, use?
• What biological self-regulatory mechanisms or rules of neural self-organization govern the networking in the human brain?
• What sociological self-regulatory mechanisms or rules of social self-organization govern the networking in a social system?
• What parallels exist from a biological and sociological perspective, and what factors influence self-organized networking in both systems?
The self-organizing mechanisms of biological and sociological systems show remarkable parallels, even though they operate at different levels—in the brain at the level of neurons and in social systems at the level of people. Both systems are based on information, communication, and interaction processes that enable them to adapt dynamically to changes and form complex networks.
A central principle at work in both neural and social networks is plasticity. The brain exhibits a remarkable ability to form new connections and adapt existing ones to enable learning, memory formation, or rehabilitation following injuries. Social systems demonstrate similar adaptability when adjusting to new technological, political, or cultural conditions—for example, through new organizational structures, altered forms of communication, or innovative social practices.
Another characteristic of both systems is the decentralization of control. The brain is not a hierarchically controlled system, but organizes itself through the interaction of many specialized regions and networks. Similarly, social systems such as markets, open-source communities, or democratic societies often develop without central control, as collective patterns emerge through the local interaction of many actors.
I think we should consider how to build I&C and AI systems to promote democratic information, communication, and interaction processes and to limit autocratic processes. Here we are again with Gary’s neurosymbolic AI—or, as I’d like to call it, a “federated neurosymbolic hybrid HCAI”!
Friedrich Schieck: I linked to your work on your Substack and did only a brief perusal of what I'll certainly have a fuller look at later; but I wanted to comment at several stopping points. It'll have to wait till later, but I think others here might also want to do a deep dive into its nuance, as I did. Impressive to me, like Gary's attitude, though my "look" was brief, I had the impression that you have not claimed anything you cannot give a reasonable account of. That's quite meaningful in today's world.
The greater "we" are in a strange place at present and I am glad to see some voices, like Gary's, being heard.
Yes!!! Yes yes yes! You are absolutely correct, Catherine, and I want to take your last line and shout it from the hilltops:
WHEN ONE PERSON CAN BLOW UP THE WORLD, WE KNOW WE HAVE DONE SOMETHING WRONG.
What this administration has done is nothing less than using the power that already existed, that America had already *allowed* to become concentrated by using the argument that "No one who successfully gets to the Presidency would ever abuse that power."
Regardless of whether and how much that may have been true in the past, the past years have been an extremely informative education of exactly how bad things can get when we let power become that concentrated. Unfucking The World is starting out by taking aim at that concentration of power itself, adding a one more to the Constitution's list of checks and balances: this will be one that allows the people to say "NO MORE" when the Federal Government no longer has the best interests of the American People at heart.
Other countries already have this: look up "vote of no confidence" on Wikipedia. The first goal of Unfucking The World is to ratify a No Confidence Amendment to the US Constitution—and if you want to know more about that, I'll invite you to subscribe to UTW (for free, I'm not accepting any money for this) and learn more on June 29!
Chris Blue: My guess is that too many (even in the USA) don't know the radical difference between democracy (as a dynamism of power in the people) and more top-down power structures, especially kingship and its derivative, authoritarianism. In the United States, the founders made a 180 turn away from that top-down source of power (written right into the Constitution) and into a formal and systematic order--which is not to say everyone automatically understands the difference or knows how to live in it.
BTW, this great difference, and its inherent break with the principles that govern families (more top-down), is why public education and cultural institutions take the center stage in keeping democracy going towards
accepting and living with the tensions inherent in democratic structures.
But the U.S. Constitution itself also reflects the flexibility of working through the inherent problems of the Constitution itself, and of course changes that occur in history that call for changing views at the dynamic intersection between "the people" and their government.
It's not "fantasy," and it may not be truthful in every case; but it IS the search for truth, and it is definitely not for political sissies.
Yeah what you’re saying is just simply not true. The founding fathers of America believed the owners of the country should run it. They cited “we the people” as debt-free property owners. Not citizens. Citizens United ruled “corporations are ‘the people’.
The U.S. is not a democracy. Not even close. You don’t comprehend the reality of what you live in. The public education system and the cultural institutions are designed to keep you stuck in a fantasy. A fantasy where you do as you’re told and fight for things that don’t exist. A collective insanity.
I suggest you read the recent book Raised to Obey. By UC professor Augustina Paglayan:
A very good question, Percy! And the answer is twofold. The first part is that this is in fact the reason why UTW isn't addressing the AI policy issue (or the human rights issues, or the economic issues) directly.
Unfucking The World is a purely non-partisan movement; it will grant no advantage to either of the Democratic or Republican parties over the other. In fact, I expect to be making quite the enemy of both major parties, as a post-UTW world will see both of them with vastly reduced political power compared to what they enjoy today. I am reaching out, not to the liberals of the US or to the conservatives, but to the *electorate*, on both sides of the aisle.
The other part of the answer is that the legislative strategy I'm using is only accessible when roughly the entirety of America feels that the existing system offers no choice of things getting better. If UTW succeeds, *or* if something else happens to give people hope (whether that hope is justified or manufactured), then it will no longer be possible to build the kind of coalition that it would take to push an amendment through this way, because people will want to wait and see if things get better on their own first.
Danielle, you seem to have a very inaccurate grip on reality. The U.S. government is working flawlessly for the people that own and control it. You’re trying something that’s been tried infinite times with 0% success.
Your argument seems to rely on flawed assumption that the citizens are in control of the country. That’s not reality. The U.S. is a top down control system not a bottom up.
“this will be one that allows the people to say "NO MORE" when the Federal Government no longer has the best interests of the American People at heart.”
The U.S. government has never had the best interests of the American people at heart. That’s an absolutely ridiculous thing to believe. I could recommend some books for you to read that could right your ship because you’ve built your foundation of knowledge on some very unstable fundamentals.
Chris Blue: I wrote my recent note to you before I read the above to Danielle:
But it appears to me that you are making your analysis on short-term real politic, which makes Trump the normative rule. (I don't think so.) . . .instead of the principles that are there for all to read in the U.S. Constitution not to mention in a good part of our concrete history over the past two+ centuries.
We can but we don't have to go back "behind" the Constitution to unearth its philosophical foundations to understand that it is of/by/for the people (180 from other forms of government) or to realize both its problems and its inspirations towards the well-being of us all. Personally, I cannot imagine someone living in this country and saying what you say about its foundations. It occurred to me to ask: are you a troll?
The reason you can’t imagine someone saying what I say about this country’s foundations is because you’ve been brainwashed by the public education and the cultural institutions that make sure you don’t comprehend reality.
It is not “short term realpolitic” it is traditional realpolitik.
As I recommended earlier you should read Raised To Obey. You don’t know that you don’t even know what’s happening.
In that case, I invite you to subscribe to UTW in order to watch me crash and burn! Perhaps you'll be surprised, or perhaps you'll just get to have the schadenfreude of watching me try and fail. You are more than welcome to say "I told you so" in the comments!
I'd offer you a gift subscription, but I'm not accepting any money for UTW, so it wouldn't do much good. So basically, what I'm saying is:
Hold my beer, Chris, and watch this. You'll have fun either way. 😁
No thanks. I don’t fill my brains limited storage with things that aren’t physically possible. I might as well listen to someone describing how I can fly if I flap my arms fast enough in a particular fashion.
I'll actually go a little further than that: the last time an amendment was both proposed AND ratified was in 1971, and that (given the ⅔ of both houses of Congress requirement for proposal) is a much more relevant figure. We're stuck at the Congress level, not at the state level.
I promise you that anything you can find with a five-minute Wikipedia search is something that I've already taken into account. You may only just be learning about UTW today, but I've been working on this ever since November.
So I'll ask you this: given that I'm not accepting any money for this, given that if I fail in my aims I will have made myself a public laughingstock for the rest of my life, given that I in fact live in the US myself, what do you believe would inspire me to dedicate so much of my own time and money and energy to this project *other* than that I truly believe it stands a chance of success?
And yes, you are absolutely correct about the size of the task I'm taking on, and the amount of powerful anger I'm tempting. There's a good reason I call my project "Unfucking The World"; that is the scale I'm operating at. (And yes, I'm taking measures to physically protect myself, don't worry.)
Here's the thing about moneyed interests. They are very good at affecting percentages in one direction or another, but they can't push any chance all the way, either to 100% or to 0%. That means that there is a chance, however slight, that I do, in fact, successfully achieve my aims before they are capable of mounting a defense strong enough to shut me down.
Personally, I've decided that "a slim chance to actually literally save the whole damn world" is worthwhile enough for me to devote a year of my life and the majority of my savings to (so that I can refuse to accept any money from subscribers and still keep the lights on while this plays out).
I'm putting everything I have and I am into UTW for the simple reason that, after all is said and done, I want to be able to look at myself in the mirror and say "I did literally everything I could to make a difference. I held nothing back."
It will either be enough to save the world, or it will be enough that I can tell myself there truly is nothing I could have done, and I'll never have to wonder if I could have done more.
Brooklyn Expat: To the MAGA people who may be reading this--nothing excites Trump more than "shooting first," but then he gets bored as he is blindsided by the truths he should have known in the first place. That's what MAGA (and now everyone else) gets when "we" give the keys to the kingdom to a guy whose self-service and vindictiveness knows no bounds.
Outside the US a Gov or Business that is planning to use a US based AI, will be having second thoughts after seeing the Commerce Dept basically turn off the lights of a running system. It also raises questions of the wider world of software as a service. Who could be sure that a capricious US Gov would not do this for any software as a service originating from the US as a tool of coercion.
You say this as it would be a bad thing ? ;-) As a European I welcome these open signs and evidence that buying US gear for anything important *really* and * evidently* carries the risk that a mentally challenged and unstable Uncle Sam has the main key to the house. A wakeup call for us once more, with a swift kick in the butt added for good measure :-)
I said it in the context of being an observer. The World is becoming multi-polar and the US over the last 18 moth has shown that it cannot be seen as a trusted partner.
This action by the Commerce Dept, has implication wider than Europe or any other country. The action of the Commerce Dept whereby foreign national within a company in the US are band from accessing this AI, any company will need to reassess their business continuity plan and CIO's will need that take on board political risk. France, has moved, they are dumping Microsoft and changing over the Linux within all Gov dept's. Whats better the AI at the cutting edge that can be cut off or an AI one has more control over?
Gary, your industry and tech critiques on AI are always prescient. However, your solutions appealing to government to step in, while perhaps the only visible remedy, seem to miss the political process we have experienced over the past 4 administrations. We’re seeing the AI industry follow in the footsteps of the cryptocurrency one, and just buy politicians’ votes. We’re seeing laws and regulations that do little more than to give lip service to issues rather than actually address or try to solve any. Look, even in this EO that you reference, it’s yelling the lip service stuff, while being toothless in its remedies. Why you keep harping on government’s role here boggles the mind.
Government, as you’ve probably noticed is not the panacea of rule-making you seem to think it is. The oversight role you conjure up, is one that would require people with an ability to truly understand the damage potential or lack thereof, of these technologies. Those who best understand, have little to no desire to be involved in the political process or run for office. They’re smarter than that 😉. While I agree with you in principle, in practice I just don’t see your reliance on government as a path to properly addressing any of these challenges. Sadly, I have no better alternative to offer, so perhaps you’re at least proposing a solution within the realm of what’s available, and i guess that’s a valiant effort 😃
That’s the real challenge, finding the adults in the room. The companies can’t be trusted, government can’t be trusted and/or doesn’t understand the tech (they understand the interests they serve ;), and as you mentioned, people are sheeple to the tech. We’re truly f-ed 🤦🏽
See, your mistake is thinking anything coming out of the White House is a "plan".
All Trump cares about is grift. Then many of his senior aides and hangers-on are there to pick up his droppings. The next tier is there for whatever power they can grab and the pay, because they're largely incompetent and untalented. And the mid-levels are true acolytes and cult members.
Almost nobody is actually working on real policy or planning. They're just proposing blanket cuts to things they don't like without any concern for the primary or secondary effects on Americans or the USA, and enjoying watching things they touch blow up - figuratively and literally.
"Almost nobody is actually working on real policy or planning" True, but such work would require a level of independent thinking and backbone to stand up against the president once in a while. Having these attributes was a sure way to *not* be included in DTs administration. The members were chosen purely on their desire to please DT and to fulfill his every wishes, no questions asked. Steve Miller and "Project 2025" members did an excellent job.
The problem is deeper than the Trump White House. For 200 years, engineers have built devices that they understand, so they can prove that those devices will not do certain kinds of harm. Neural nets are not like that. Nobody really understands how they work. There is no tractable theory of how a neural net works - so nobody can prove that it will not do harm X or harm Y. Not by testing it, or by formal proof, or in court. You can always jail-break an LLM, if there is no world model to stop you.
I follow you regularly. Today I realized you used many images/screenshots, and on email, I couldn't see alternative text on them. This makes it inaccessible for someone using a screenreader. I hope you will fix this so I can share with my blind friends. (AI can create draft alternative text to help you get started!)
Technically well-informed governance? Hahahaha, ahhhhh, tell me another one, Gary.
Caveat emptor and use at everyone else's risk has been the model in software long before LLMs came along. Somehow we just accepted "that would be too hard" and "I don't wanna" as reasonable excuses for buggy, insecure, and unreliable software, as long as somewhere at the bottom of the unilateral license agreement that nobody reads one put the sacred disclaimer against liability boilerplate. Which is one thing when that software is somebody's hobby project, another thing when it's flying airplanes, making medical decisions, regulating infrastructure, or driving cars.
Red flags should have been raised the moment anybody shoved some "don't blame me, I didn't feel like testing it" code anywhere where lives were at stake. Instead we enthusiastically embraced doing so in everything from pacemakers to toasters. Now here we are, eating our just desserts, with the ultimate in eats-the-world shitware. Oh well. Mistakes were made. Hindsight is 20/20.
Real question is why going forward any company anywhere outside the US would ever rely on a US AI model for any significant task, knowing that access could be withdrawn by a US government order at any moment. This is the way for the US to compete effectively in the global AI market??
"thriving industry"? - I think you are mistaking hype and the foolishness of those taken in by the hype for thriving. Just because it is fruit does not make it good fruit. Just because people prophesise does not make them prophets.
It is true that a lot of people are making a lot of money. All those billions are being spent. I myself used to be a translator until AI took all the work, and then I made a reasonable living correcting AI output for the training/algorithms. Now even that has dried up. So money is circulating, just like military spending during military conflict, but I couldn't say that that money couldn't be better spent elsewhere. And when people do realise that it is not intelligence but only seems like it, hopefully that money will go somewhere more productive.
It’s hard to imagine anything LESS productive than a system that uses as much capital, energy and fresh water as LLMs to produce unreliable answers that must be independently checked and verified in every case.
Unless maybe you are talking about the universe itself when it it undergoing its ultimate heat death.
"All those billions are being spent" There is quite some doubt as the financial circle jerk, where most of these billions are publicly mentioned, seems more of a spreadsheet excercise than real money changing hands. But who knows?
"hopefully that money will go somewhere more productive." Amen to that!
I somehow missed the mention of what the Commerce Department export control directive actually says, and the link to it. Maybe someone could direct to me where that is in the article.
You know it’s almost as if the Trump admin was totally clueless and focused only on culture wars issues rather than grown-up governing.
Yes, almost...
It's the woke gang
Their base wants it that way.
Come on! Governing is for sissies! We are manly machos! (Or was that nachos? I forget, I did not pay attention at school).
You have it spot-on as always, Gary. If you find it as unlikely as I do that our current government can manage any of your three criteria, you might be interested to know that the newsletter/podcast/movement I'll be launching on June 29 will provide a way for the people to demand a government that can make good decisions for us, if enough of us speak with one voice.
say more?
I'll be launching a newsletter/podcast/political movement called Unfucking The World, with the explicit goal to use a novel legislative strategy to force an amendment through to our Constitution in less time than anyone imagined possible.
The amendment is drafted, the strategy is set, (most of) the podcast episodes are recorded, and my savings are sufficient to see the project through to any successful conclusion: it'll either work within a year, or it will have lost momentum and failed. (I'm not accepting any money from anyone, to avoid accusations of scam.)
UTW isn't going to address AI policy on its own: it will give us a government that we can trust to make decisions about AI policy for us, and one that (I imagine) will be much more likely to take your concerns seriously.
Danielle Church: A related narrative:
The current situation suggests that we need to name a REAL distinction that gets blurred when we are discussing ideas about "the government."
There is the drift in thinking of "the government" as (a) a singular ideology aimed at having more and more power, and leaning towards unmitigated USE of and control of The People," e.g., the unitary executive authority idea fostered by the Heritage Foundation and its control-freaks; . . .
. . . and (b) "the government" (referring to a democracy in the case of the United States) thought of as (recovered as meant originally) a dynamic, working, civilized, intelligent, and reasonable set of true representatives of The People openly working through these problems for the good of all concerned.
The first interpretation of "the government" we are and should be afraid of, including (and unquestionably) a tendency towards PRIVATE economics holding all the power cards; the second we need to understand ourselves and everyone else as a working part of what needs to remain DYNAMIC and interactive, and more on the side of ideas of the dynamism of what needs to remain in the PUBLIC DOMAIN, including of the economics that underpin everyone's well-being.
When one person can blow up the world, we know we have done something wrong.
Every social system (whether a family, a company, or a state) is based on—or develops and changes based on—the information, communication, and interaction processes of the individuals involved. The purpose of digital information and communication (I&C) and AI systems is precisely to automate, monitor, and control these processes, thereby steering a social system in one direction or another.
One direction could be a democratic system; the other could be an autocratic system or a dictatorship. The fundamental question this raises is: How must I&C and AI systems be designed to steer a social system in one direction or another?
I would phrase the question differently: What rules or self-regulatory mechanisms do natural systems, for example, use?
• What biological self-regulatory mechanisms or rules of neural self-organization govern the networking in the human brain?
• What sociological self-regulatory mechanisms or rules of social self-organization govern the networking in a social system?
• What parallels exist from a biological and sociological perspective, and what factors influence self-organized networking in both systems?
The self-organizing mechanisms of biological and sociological systems show remarkable parallels, even though they operate at different levels—in the brain at the level of neurons and in social systems at the level of people. Both systems are based on information, communication, and interaction processes that enable them to adapt dynamically to changes and form complex networks.
A central principle at work in both neural and social networks is plasticity. The brain exhibits a remarkable ability to form new connections and adapt existing ones to enable learning, memory formation, or rehabilitation following injuries. Social systems demonstrate similar adaptability when adjusting to new technological, political, or cultural conditions—for example, through new organizational structures, altered forms of communication, or innovative social practices.
Another characteristic of both systems is the decentralization of control. The brain is not a hierarchically controlled system, but organizes itself through the interaction of many specialized regions and networks. Similarly, social systems such as markets, open-source communities, or democratic societies often develop without central control, as collective patterns emerge through the local interaction of many actors.
I think we should consider how to build I&C and AI systems to promote democratic information, communication, and interaction processes and to limit autocratic processes. Here we are again with Gary’s neurosymbolic AI—or, as I’d like to call it, a “federated neurosymbolic hybrid HCAI”!
https://friedrichschieck.substack.com/p/hybrid-hcai-a-thought-experiment,
Friedrich Schieck: I linked to your work on your Substack and did only a brief perusal of what I'll certainly have a fuller look at later; but I wanted to comment at several stopping points. It'll have to wait till later, but I think others here might also want to do a deep dive into its nuance, as I did. Impressive to me, like Gary's attitude, though my "look" was brief, I had the impression that you have not claimed anything you cannot give a reasonable account of. That's quite meaningful in today's world.
The greater "we" are in a strange place at present and I am glad to see some voices, like Gary's, being heard.
https://friedrichschieck.substack.com/p/from-bcm-to-federated-neuro-symbolic
Yes!!! Yes yes yes! You are absolutely correct, Catherine, and I want to take your last line and shout it from the hilltops:
WHEN ONE PERSON CAN BLOW UP THE WORLD, WE KNOW WE HAVE DONE SOMETHING WRONG.
What this administration has done is nothing less than using the power that already existed, that America had already *allowed* to become concentrated by using the argument that "No one who successfully gets to the Presidency would ever abuse that power."
Regardless of whether and how much that may have been true in the past, the past years have been an extremely informative education of exactly how bad things can get when we let power become that concentrated. Unfucking The World is starting out by taking aim at that concentration of power itself, adding a one more to the Constitution's list of checks and balances: this will be one that allows the people to say "NO MORE" when the Federal Government no longer has the best interests of the American People at heart.
Other countries already have this: look up "vote of no confidence" on Wikipedia. The first goal of Unfucking The World is to ratify a No Confidence Amendment to the US Constitution—and if you want to know more about that, I'll invite you to subscribe to UTW (for free, I'm not accepting any money for this) and learn more on June 29!
The U.S. has always been (a) and it has never been (b).
You could better describe the U.S. system as
(a) truth
(b) fantasy
Chris Blue: My guess is that too many (even in the USA) don't know the radical difference between democracy (as a dynamism of power in the people) and more top-down power structures, especially kingship and its derivative, authoritarianism. In the United States, the founders made a 180 turn away from that top-down source of power (written right into the Constitution) and into a formal and systematic order--which is not to say everyone automatically understands the difference or knows how to live in it.
BTW, this great difference, and its inherent break with the principles that govern families (more top-down), is why public education and cultural institutions take the center stage in keeping democracy going towards
accepting and living with the tensions inherent in democratic structures.
But the U.S. Constitution itself also reflects the flexibility of working through the inherent problems of the Constitution itself, and of course changes that occur in history that call for changing views at the dynamic intersection between "the people" and their government.
It's not "fantasy," and it may not be truthful in every case; but it IS the search for truth, and it is definitely not for political sissies.
Yeah what you’re saying is just simply not true. The founding fathers of America believed the owners of the country should run it. They cited “we the people” as debt-free property owners. Not citizens. Citizens United ruled “corporations are ‘the people’.
The U.S. is not a democracy. Not even close. You don’t comprehend the reality of what you live in. The public education system and the cultural institutions are designed to keep you stuck in a fantasy. A fantasy where you do as you’re told and fight for things that don’t exist. A collective insanity.
I suggest you read the recent book Raised to Obey. By UC professor Augustina Paglayan:
https://a.co/d/0brXSKnf
I’d also recommend How The World Works by Noam Chomsky.
One question. Couldn't conservatives use the same legal strategy to immediately repeal your amendment whatever it may be?
A very good question, Percy! And the answer is twofold. The first part is that this is in fact the reason why UTW isn't addressing the AI policy issue (or the human rights issues, or the economic issues) directly.
Unfucking The World is a purely non-partisan movement; it will grant no advantage to either of the Democratic or Republican parties over the other. In fact, I expect to be making quite the enemy of both major parties, as a post-UTW world will see both of them with vastly reduced political power compared to what they enjoy today. I am reaching out, not to the liberals of the US or to the conservatives, but to the *electorate*, on both sides of the aisle.
The other part of the answer is that the legislative strategy I'm using is only accessible when roughly the entirety of America feels that the existing system offers no choice of things getting better. If UTW succeeds, *or* if something else happens to give people hope (whether that hope is justified or manufactured), then it will no longer be possible to build the kind of coalition that it would take to push an amendment through this way, because people will want to wait and see if things get better on their own first.
Danielle, you seem to have a very inaccurate grip on reality. The U.S. government is working flawlessly for the people that own and control it. You’re trying something that’s been tried infinite times with 0% success.
Your argument seems to rely on flawed assumption that the citizens are in control of the country. That’s not reality. The U.S. is a top down control system not a bottom up.
“this will be one that allows the people to say "NO MORE" when the Federal Government no longer has the best interests of the American People at heart.”
The U.S. government has never had the best interests of the American people at heart. That’s an absolutely ridiculous thing to believe. I could recommend some books for you to read that could right your ship because you’ve built your foundation of knowledge on some very unstable fundamentals.
Chris Blue: I wrote my recent note to you before I read the above to Danielle:
But it appears to me that you are making your analysis on short-term real politic, which makes Trump the normative rule. (I don't think so.) . . .instead of the principles that are there for all to read in the U.S. Constitution not to mention in a good part of our concrete history over the past two+ centuries.
We can but we don't have to go back "behind" the Constitution to unearth its philosophical foundations to understand that it is of/by/for the people (180 from other forms of government) or to realize both its problems and its inspirations towards the well-being of us all. Personally, I cannot imagine someone living in this country and saying what you say about its foundations. It occurred to me to ask: are you a troll?
The reason you can’t imagine someone saying what I say about this country’s foundations is because you’ve been brainwashed by the public education and the cultural institutions that make sure you don’t comprehend reality.
It is not “short term realpolitic” it is traditional realpolitik.
As I recommended earlier you should read Raised To Obey. You don’t know that you don’t even know what’s happening.
In that case, I invite you to subscribe to UTW in order to watch me crash and burn! Perhaps you'll be surprised, or perhaps you'll just get to have the schadenfreude of watching me try and fail. You are more than welcome to say "I told you so" in the comments!
I'd offer you a gift subscription, but I'm not accepting any money for UTW, so it wouldn't do much good. So basically, what I'm saying is:
Hold my beer, Chris, and watch this. You'll have fun either way. 😁
No thanks. I don’t fill my brains limited storage with things that aren’t physically possible. I might as well listen to someone describing how I can fly if I flap my arms fast enough in a particular fashion.
You are correct! They wanted to make sure it wouldn't ever happen unless the country REALLY wanted it.
I'll actually go a little further than that: the last time an amendment was both proposed AND ratified was in 1971, and that (given the ⅔ of both houses of Congress requirement for proposal) is a much more relevant figure. We're stuck at the Congress level, not at the state level.
I promise you that anything you can find with a five-minute Wikipedia search is something that I've already taken into account. You may only just be learning about UTW today, but I've been working on this ever since November.
So I'll ask you this: given that I'm not accepting any money for this, given that if I fail in my aims I will have made myself a public laughingstock for the rest of my life, given that I in fact live in the US myself, what do you believe would inspire me to dedicate so much of my own time and money and energy to this project *other* than that I truly believe it stands a chance of success?
All I can say is "Good luck with that"...
You are taking on the entire moneyed interests of this country and beyond. You are going against the ENTIRE history of the state.
I'd say, "Watch your back" is the least you should do.
Thank you for the wishes!
And yes, you are absolutely correct about the size of the task I'm taking on, and the amount of powerful anger I'm tempting. There's a good reason I call my project "Unfucking The World"; that is the scale I'm operating at. (And yes, I'm taking measures to physically protect myself, don't worry.)
Here's the thing about moneyed interests. They are very good at affecting percentages in one direction or another, but they can't push any chance all the way, either to 100% or to 0%. That means that there is a chance, however slight, that I do, in fact, successfully achieve my aims before they are capable of mounting a defense strong enough to shut me down.
Personally, I've decided that "a slim chance to actually literally save the whole damn world" is worthwhile enough for me to devote a year of my life and the majority of my savings to (so that I can refuse to accept any money from subscribers and still keep the lights on while this plays out).
I'm putting everything I have and I am into UTW for the simple reason that, after all is said and done, I want to be able to look at myself in the mirror and say "I did literally everything I could to make a difference. I held nothing back."
It will either be enough to save the world, or it will be enough that I can tell myself there truly is nothing I could have done, and I'll never have to wonder if I could have done more.
“Shoot first and ask questions later” explains way too much USG policy at the moment, both literally and figuratively.
Brooklyn Expat: To the MAGA people who may be reading this--nothing excites Trump more than "shooting first," but then he gets bored as he is blindsided by the truths he should have known in the first place. That's what MAGA (and now everyone else) gets when "we" give the keys to the kingdom to a guy whose self-service and vindictiveness knows no bounds.
It IS the current govt. regime's policy, along with the AI tech titans' policy.
"Questions ? What questions? Why? What are you blabbering about ?" would sum up the current administration's stance a bit better.
Outside the US a Gov or Business that is planning to use a US based AI, will be having second thoughts after seeing the Commerce Dept basically turn off the lights of a running system. It also raises questions of the wider world of software as a service. Who could be sure that a capricious US Gov would not do this for any software as a service originating from the US as a tool of coercion.
You say this as it would be a bad thing ? ;-) As a European I welcome these open signs and evidence that buying US gear for anything important *really* and * evidently* carries the risk that a mentally challenged and unstable Uncle Sam has the main key to the house. A wakeup call for us once more, with a swift kick in the butt added for good measure :-)
I said it in the context of being an observer. The World is becoming multi-polar and the US over the last 18 moth has shown that it cannot be seen as a trusted partner.
This action by the Commerce Dept, has implication wider than Europe or any other country. The action of the Commerce Dept whereby foreign national within a company in the US are band from accessing this AI, any company will need to reassess their business continuity plan and CIO's will need that take on board political risk. France, has moved, they are dumping Microsoft and changing over the Linux within all Gov dept's. Whats better the AI at the cutting edge that can be cut off or an AI one has more control over?
Gary, your industry and tech critiques on AI are always prescient. However, your solutions appealing to government to step in, while perhaps the only visible remedy, seem to miss the political process we have experienced over the past 4 administrations. We’re seeing the AI industry follow in the footsteps of the cryptocurrency one, and just buy politicians’ votes. We’re seeing laws and regulations that do little more than to give lip service to issues rather than actually address or try to solve any. Look, even in this EO that you reference, it’s yelling the lip service stuff, while being toothless in its remedies. Why you keep harping on government’s role here boggles the mind.
Government, as you’ve probably noticed is not the panacea of rule-making you seem to think it is. The oversight role you conjure up, is one that would require people with an ability to truly understand the damage potential or lack thereof, of these technologies. Those who best understand, have little to no desire to be involved in the political process or run for office. They’re smarter than that 😉. While I agree with you in principle, in practice I just don’t see your reliance on government as a path to properly addressing any of these challenges. Sadly, I have no better alternative to offer, so perhaps you’re at least proposing a solution within the realm of what’s available, and i guess that’s a valiant effort 😃
If not "government" then who? "The people"? who can't stop scrolling Instagram?
That’s the real challenge, finding the adults in the room. The companies can’t be trusted, government can’t be trusted and/or doesn’t understand the tech (they understand the interests they serve ;), and as you mentioned, people are sheeple to the tech. We’re truly f-ed 🤦🏽
See, your mistake is thinking anything coming out of the White House is a "plan".
All Trump cares about is grift. Then many of his senior aides and hangers-on are there to pick up his droppings. The next tier is there for whatever power they can grab and the pay, because they're largely incompetent and untalented. And the mid-levels are true acolytes and cult members.
Almost nobody is actually working on real policy or planning. They're just proposing blanket cuts to things they don't like without any concern for the primary or secondary effects on Americans or the USA, and enjoying watching things they touch blow up - figuratively and literally.
"Almost nobody is actually working on real policy or planning" True, but such work would require a level of independent thinking and backbone to stand up against the president once in a while. Having these attributes was a sure way to *not* be included in DTs administration. The members were chosen purely on their desire to please DT and to fulfill his every wishes, no questions asked. Steve Miller and "Project 2025" members did an excellent job.
First, how did Florida's interest in helping the people get past Trump's "Let's Save All the Criminals" working policy?
Did the order come from Donnie or from Elon?
I'm guessing B.
The problem is deeper than the Trump White House. For 200 years, engineers have built devices that they understand, so they can prove that those devices will not do certain kinds of harm. Neural nets are not like that. Nobody really understands how they work. There is no tractable theory of how a neural net works - so nobody can prove that it will not do harm X or harm Y. Not by testing it, or by formal proof, or in court. You can always jail-break an LLM, if there is no world model to stop you.
I follow you regularly. Today I realized you used many images/screenshots, and on email, I couldn't see alternative text on them. This makes it inaccessible for someone using a screenreader. I hope you will fix this so I can share with my blind friends. (AI can create draft alternative text to help you get started!)
Technically well-informed governance? Hahahaha, ahhhhh, tell me another one, Gary.
Caveat emptor and use at everyone else's risk has been the model in software long before LLMs came along. Somehow we just accepted "that would be too hard" and "I don't wanna" as reasonable excuses for buggy, insecure, and unreliable software, as long as somewhere at the bottom of the unilateral license agreement that nobody reads one put the sacred disclaimer against liability boilerplate. Which is one thing when that software is somebody's hobby project, another thing when it's flying airplanes, making medical decisions, regulating infrastructure, or driving cars.
Red flags should have been raised the moment anybody shoved some "don't blame me, I didn't feel like testing it" code anywhere where lives were at stake. Instead we enthusiastically embraced doing so in everything from pacemakers to toasters. Now here we are, eating our just desserts, with the ultimate in eats-the-world shitware. Oh well. Mistakes were made. Hindsight is 20/20.
Fukitol: AKA, writings on the walls of a tomb have no meaning if there is no one left to read them.
Lobbyists control the government.
Lobbyists bribe the politicians.
Lobbyists even write the legislation&give it to the politicians.
Real question is why going forward any company anywhere outside the US would ever rely on a US AI model for any significant task, knowing that access could be withdrawn by a US government order at any moment. This is the way for the US to compete effectively in the global AI market??
"thriving industry"? - I think you are mistaking hype and the foolishness of those taken in by the hype for thriving. Just because it is fruit does not make it good fruit. Just because people prophesise does not make them prophets.
It doesn’t even make them profits.
Aren’t those supposed to be part of a “thriving industry”?
Quite.
It is true that a lot of people are making a lot of money. All those billions are being spent. I myself used to be a translator until AI took all the work, and then I made a reasonable living correcting AI output for the training/algorithms. Now even that has dried up. So money is circulating, just like military spending during military conflict, but I couldn't say that that money couldn't be better spent elsewhere. And when people do realise that it is not intelligence but only seems like it, hopefully that money will go somewhere more productive.
It’s hard to imagine anything LESS productive than a system that uses as much capital, energy and fresh water as LLMs to produce unreliable answers that must be independently checked and verified in every case.
Unless maybe you are talking about the universe itself when it it undergoing its ultimate heat death.
"All those billions are being spent" There is quite some doubt as the financial circle jerk, where most of these billions are publicly mentioned, seems more of a spreadsheet excercise than real money changing hands. But who knows?
"hopefully that money will go somewhere more productive." Amen to that!
It’s certainly the only industry in the history of the world for which losing money hand over first actually boosts company value.
I somehow missed the mention of what the Commerce Department export control directive actually says, and the link to it. Maybe someone could direct to me where that is in the article.
see yesterday’s essay
I'm still not sure this was anything but a marketing ploy. Maybe I've just gone around the bend. That's a real risk with the times we're in.