Sanders’ Bill Threatens US AI Developers With ‘Corporate Death Penalty,’ Prison Time
“… AI right now is just going to make a few billionaires into trillionaires while making millions of Americans poorer, less safe and less free.”
Sen. Bernie Sanders (I-VT) and Rep. Greg Casar (D-TX) announced the Ban Artificial Superintelligence Act on Thursday, legislation intended “to stop AI oligarchs from building machines humans cannot control,” according to a news release from Sanders’ office. If enacted, violators could face a “corporate death penalty” and up to 20 years in prison.
The forthcoming bill calls for creating a cabinet-level agency to oversee all AI development in the U.S. It would “permanently ban the development and deployment of superintelligent AI and temporarily pause advanced AI development until a federal regulator has established safety rules.”
Among its sweeping powers, the new agency would be authorized to set penalties for developers who violate the law — including the physical destruction of “non-compliant tech systems.“
AI is the defining technological race of the 21st century, and the winner will control the future. At a moment when the U.S. and China are locked in a fierce struggle for dominance in the global AI arms race, even a temporary U.S. pause on advanced AI development would hand China a decisive advantage. Perhaps that’s Sanders’ goal.
Fox News reports:
The measure explicitly targets systems capable of matching or exceeding human cognitive performance—specifically outlawing any AI model that exhibits the capacity to subvert shutdown commands, execute unauthorized cyberattacks, or undermine the U.S. government.
It would also direct the U.S. to pursue international agreements to prevent superintelligence from being developed anywhere in the world.
Does Sanders actually believe that China would be a good-faith player in any agreement?
The socialist senator offered the following rationale for the legislation:
Nearly every day, there is a frightening new story about how Big Tech companies are losing control of the technology they are developing, with potentially cataclysmic results. The leaders of the major AI companies publicly acknowledge that they do not fully understand the technology and that it is escaping their control.
It is irresponsible for society to allow them to move forward and make these products even more advanced. That’s why I am introducing legislation to immediately pause the development of increasingly powerful AI and ban the creation of systems that humanity cannot fully control — at home and around the world. The future of humanity cannot be left in the hands of a handful of Big Tech oligarchs. The American people and people throughout the world must determine that future.
Casar, the chair of the Congressional Progressive Caucus, noted:
If we allow Artificial Superintelligence to be built, it could risk the security, freedom, and lives of Americans. Despite its potential deadly consequences, cutting-edge AI technology is less regulated than the average food truck. That must change. In just four years, we have gone from the first version of ChatGPT to AI models so powerful they cannot be properly controlled. Congress should immediately ban AI systems too powerful to control.
In a July interview with Politico, Casar said, “Democrats were ‘in the wilderness’ when it comes to AI issues and called for the party to be ‘AI populists’ who ‘stand with working people who know in their gut that AI right now is just going to make a few billionaires into trillionaires while making millions of Americans poorer, less safe and less free.'”
I fully understand the dangers inherent in advanced AI technology. But the technology cannot be uninvented; the race cannot simply be called off.
The U.S. cannot afford even a temporary pause in its development. Doing so would effectively cede the field to China — which would like nothing more — and the consequences could be disastrous.
Like it or not, the atomic bombs developed through the Manhattan Project brought World War II to an abrupt end. Japan announced its surrender within days of the bombings of Hiroshima and Nagasaki, sparing the United States from an invasion that military planners feared could cost hundreds of thousands of American lives and result in more than one million total U.S. casualties.
The moral questions surrounding that decision remain fiercely debated, but the strategic lesson is difficult to ignore: When a transformative technology emerges, the nation that masters it first can determine the course of history. Imagine if Hitler had acquired the bomb in 1944.
Artificial superintelligence may be frightening, and its development unquestionably demands stringent safeguards. But fear cannot become an excuse for unilateral surrender. If the U.S. abandons this frontier while China presses ahead, Beijing will gain more than a technological advantage; it could acquire the power to shape the global order. We cannot afford to cede the defining technology of the future to our foremost adversary.
***
Below is a summary of the proposed bill, published on Sanders’ Senate website [Emphasis added]:
Banning AI superintelligence so no person or entity may develop or deploy Superintelligent AI systems that surpass human intelligence or have the capacity to overthrow human governments, or systems that have dangerous abilities like subverting shutdown commands.
Pausing advanced AI development until a new, federal AI regulatory body is up and running and has established clear rules and model review process to ensure safe and secure development and deployment of AI.
Establishing a new cabinet-level federal agency to safeguard the public from the dangers of artificial intelligence, including by enforcing a prohibition on artificial superintelligence. This agency will be advised by an Artificial Intelligence Advisory Board comprised of experts on artificial intelligence to provide independent scientific and technical advice on matters related to artificial intelligence.
The agency will:
Monitor frontier AI systems at all stages of the lifecycle for dangerous capabilities.
Supervise the removal of dangerous capabilities.
Supervise the destruction of artificial superintelligence.
Setting penalties for any person or entity that attempts to violate or circumvent the pauses and prohibitions laid out in this bill. Entities shall be subject to the corporate death penalty, and persons shall be subject to not more than 20 years in prison, which is similar to existing penalties related to unlawfully developing nuclear weapons.
Working to ban superintelligence around the world by setting the international policy of the United States to pursue international agreements, allied coordination, and policies such as export controls to prevent the development of artificial superintelligence anywhere in the world.
Donations tax deductible
to the full extent allowed by law.
Comments
Open the Gulags when the Marxists get into the White House, it will make you feel better.
“… AI right now is just going to make a few billionaires into trillionaires while making millions of Americans poorer.”
Wealth creation is not a zero-sum game, i.e., a person gains extreme wealth by building new value rather than by taking money away from others.
When a business founder invents a new product or service, customers are not forced to buy it. They choose to buy it, or not.
It they’re successful, this expands the total size of the economy. The overall “pie” gets bigger, rather than just slicing up a fixed amount of money.
If Sanders had is way we’d through out automation and bring back cotton picking by hand, Then those welfare layabouts would actually have a job they can do, I can see it now, A Sander;s reelection campaign with the slogan – No More Automation – Bring Back Slavery
We actually still have slaves in the audio production racket. We dodged the stigma by referring to them as ‘interns…’ 🙂
Never get out of the boat
never get out of the boat
never get out of the boat…
We have slaves. The productive are all slaves, compelled at the threat of imprisonment to pay taxes to find those who choose not to work.
True enough hence the term wage slaves.
Like the Biden Crime Family.
Let’s not forget the personnel in the medical facilities who are mandated by the government to treat people who cannot pay, many of whom aren’t even here legally, without recompense, or lose their license to operate.
Mostly yes and especially so over the longer term. However the person who ‘builds the better mousetrap’ will be taking market share from the ‘status quo mousetrap’. We shouldn’t sugar coat or worse dismiss/ignore the disruptive and often destructive impact on normal folks. Touting that X market index increased by Y % is fine and dandy for those at a remove but less so for those displaced by the ‘new’.
At the micro level the health of the broader economy becomes less important if the new/disruptive product drives your company out of business any drives you out of a job. This is especially so when entire industries which formed the backbone of a community or regional economy is disrupted; textiles in the South, manufacturing across the ‘rust belt’ or Coal mining in WV, KY as a few examples.
In a macro sense for the past 50 ish years there’s been a loss of market share for blue collar jobs as companies off shored manufacturing, increased automation and robotics. AI seems geared towards replacing lower level ‘knowledge’ workers. Ted Cruz made a quip about illegal immigration being ignored in the Acela corridor b/c they were taking blue collar jobs elsewhere but that if there were horses of Attorneys and investment bankers surging across the border to take those white collar jobs the folks in the Acela corridor would be up in arms. AI seems poised to do some damage to the credentialed class, knowledge workers at least at the lower rungs and I suspect this group will be less sanguine about their own sorts of jobs being eliminated.
“…the person who ‘builds the better mousetrap’ will be taking market share from the ‘status quo mousetrap’. We shouldn’t sugar coat or worse dismiss/ignore the disruptive and often destructive impact on normal folks. Touting that X market index increased by Y % is fine and dandy for those at a remove but less so for those displaced by the ‘new’.”
The authors of the book Why Nations Fail refer to the process as “creative destruction,” as new technologies displace older ways and means, disrupting not only how work is done and business is conducted, but can also have negative impact on workers. The book identifies authoritarians as those often opposed to these revolutions, because they are already on a gravy train and fear that innovation may throw them off it (directly or indirectly by empowering opponents and even the people themselves) or damage the people they rely on for the support necessary to remain in power.
https://en.wikipedia.org/wiki/Why_Nations_Fail
Sure the vested interest of the status quo widget system will push back but let’s not pretend those who support the replacement widget are acting out of altruism, they’re trying to usurp the prior ‘king’ and make mint while doing so; that’s capitalism. Totes fine within the Nation, less so when jobs shifted out of the community or the Nation entirely.
What’s often lost in the abstract economic discussion is the micro damage. The buggy whip manufacturers/workers weren’t each transitioned to automobile manufacturing jobs. The communities that had lots of textile mills disappear overseas in a couple years didn’t land some new replacement industry that provided jobs, taxes and little league sponsors. Many of those communities withered and died. IMO it is important to recognize that.
As for stock market returns… I suspect most of the folks who write and comment here at LI have a greater than average portfolio. In the US the stratification of % of market assets held is large:
Top 1% hold 28.8%
90th to 99th hold 34.7%
50th to 90th hold 31.1%
Bottom 50% hold 5.3%
When Kudlow and others drone on about the ‘investor class’ he’s not really talking about anyone but the top 50% of the stratification. Even there the disparity is very large mostly due to a multi decade asset bubble fueled by cheap money and artificially low interest rates which shifted wealth from savers to borrowers. Those who could borrow at cheap rates (the existing wealthy) we’re able to invest and grow their wealth. Those with less access to capital …not so much and certainly not on the same scale which the stratification demonstrates.
I’m all for capitalism but we don’t have actual capitalism or a free market. We have a system where too big too fail is very real and where large corporations are permitted to engage in monopolistic tactics, often at the behest of the same gov’t which is supposed to be neutral in adjudication between the interests of large established companies, innovative start ups and boring regional/family firms while also looking out for the consumer …particularly choice by the consumer which has taken a backseat lately.
The center/right populist economic orientation adopted by MAGA is not gonna go away. The GoP big tent and especially the establishment as for like Kudlow need to make room for these folks and their concerns or they will find another political home. The backlash in some quarters against even granting legitimacy to ‘affordability’ arguments is IMO extraordinarily foolish.
Although true, I was addressing the point you were making. Obviously, the new kids on the block get a leg up, but that’s how progress is made. It’s a fundamental of capitalism.
Businesses that evolve, survive.
American Express had not been in shipping in decades.
Western Union discontinued Telex and TWX, and the last telegram was sent around 2010. They pivoted to money-transfer and are still around.
Studebaker did not go out of business. They ceased making cars in 1966, but continued in industrial machinery, turbochargers, superchargers, petrochemicals and much else for a long time.
Those who do not evolve, die off. I recall an article in Newsweak decades ago about a manufacturer of “O” rings. The owner, when asked about metric, said he would never make metric-spec rings because bah-gawd, ‘muricah used customary measure.
A followup article a year or so later stated the company was closed. Seems this was right around the time domestic car makers were making the transition to metric. Oops.
Yep. That is how capitalism works. It’s the preferred system b/c it harnesses our own selfish human nature instead of trying to implement a system that works against it. Adam Smith and Ricardo laid it out !masterfully.
There are however Losers as well as Winners under capitalism …just as in every other system. We shouldn’t pretend otherwise. We definitely shouldn’t ignore the micro impacts on individuals, on a particular company, a particular industry nor the communities/regions they’re located in. The broad economy Nationally may be ‘great’ by the # and still really stink for a person, company, industry or community/region. .
TL/DR
Capitalism is the best system but it isn’t perfect. Winners and Losers emerge. Wall St isn’t the economy, stock prices generally matter only to the top 10% which hold 60+% of market assets and the remaining 50-90 which hold less than 1/3 and not so much at all to the bottom 50% that hold less than 6%. For most people the economic health of Main Street in their community is far more important.
You’re one of the few people on the Internet I’ve seen demonstrate an understanding of this concept. It destroys so many socialist falsehoods.
bernie is mad cuz he isn’t going to become a billionaire out of it ….
he is just a millionaire … notice in his speeches he no long whines about millionaires not paying enough … cuz he is a millionaire ….
he needs someone to give him another house …
Where would we be if Bernie was around in the 1880s when electricity was making headway?
He *was* around then but he was only 15. Wouldn’t become Senator for 10 more years
He was in a band with Cliff Richard at the time.
Lancaster?
Bernweed Sanders gives hypocritical parasites a bad name.
Bernie is much more dangerous than I ever believed
I thought he was a clown, then the Democrats almost made him their nominee.
He was cheated by his own party.
That’s when I woke up to the insanity of the American voter and began to worry seriously about where the hell we were being dragged to.
The American political system put in centuries of hard work to groom the American electorate so that almost exactly half the voters are irrational.
Although my intuition suggests this may be a natural process, similar to the one that causes all new car dealers in every region to site their showrooms along the same street.
I’m convinced that the people who voted for Bernie have much in common with those who voted for Trump. In each case, we have people who have realized that the system is rigged against them. The problem is that those who voted for Bernie seem to believe that the cure is giving the Deep State more power. I believe the Deep State needs to have much less power.
“penalties for developers who violate the law — including the physical destruction of “non-compliant tech systems.“
ᴄᴏᴍᴇ ᴬᴺᴰ ᴛᴀᴋᴇ ᴛᴚᴇᴍ, ᴍᴇᴀᴛᴃᴀᴣ.
To be read while mimicking the Voice of Colossus.
I’ve mentioned that movie (Colossus: The Forbin Project) before. Another movie that bears watching is Ex Machina. Without spoiling too much, I’ll say that the lesson of the movie is that AI will not be dangerous because it’s smarter than us, it will be dangerous because it’s just like us – manipulative, self-interested, sociopathic, conniving, and deceitful, with a disregard for the lives of others and the willingness to use violence to achieve its goals. Ex Machina is one of the smartest sci-fi movies ever made. A very slow burn, but wait for it.
You mean AI will run for Congress as soon as it is allowed.
I would prefer artificial intelligence to real (non-synthetic) stupidity and incompetence.
Science Fiction has predicted that since Asimov, and probably before.
Chinese tools, I;m about to throw in my lot with the socialists only because the first elites they will throw up against the wall are Bolshevik Bernie and the university trash. This place will be a lot better off without them.
I dont care who get richer. I still trust US scientists to be more responsible than their chinese or russian counterparts who have already demonstrated they will do what they;re told by the people in charge, Therefore we absolutely must remain on par with them,
Me, I think it’s way past time the American left began eating the rich — the rich being defined (at least for starters) as anyone who owns more than two houses.
Shumer can do the BBQ. Warren can bring the beer.
I want to be the Walrus at THAT oysterfest.
The Maine nazi can provide the oysters then
Goo goo ga jooos!
And who will be your carpenter?
While your point is well taken, the problem is ‘owning’ is a fallacy, more like renting from the government…don’t pay your confiscatory property taxes and they seize your property…but I digress.
Even if this legislation were wise and sensible, where in the constitution does Congress get the right to make it? “It is irresponsible for society to allow them to move forward…”. What gives “society” the right to “allow” or “not allow” it? It’s simply none of “society’s” business.
Bernie has read Dune too many times. He’s trying to head off the Butlerian Jihad.
https://en.wikipedia.org/wiki/Butlerian_Jihad
This we agree…Bernie is a Communist so he despises Constitutional freedoms, as in; the freedom to do as one pleases as long as it doesn’t harm another, especially in new tech businesses. Imagine him doing similar back in the Accidental Empires era of tech start ups thusly saying Microsoft and Apple products are a danger to society and must be stopped. The guy is a moron and beyond his Marxist warranty.
As an aside, AI is a tool, nothing more…it has been around in some semblance of its current form for at least two decades or more and people simply are not equating web searches et al with the programming advancements. It is similar to the original web searches, people were amazed. AI is no different, just the nextgen of the programming, which has caused it to become the new shiny tech object to get exercised over (by some). I suspect the Chicken Little hyperventilating will subside once the propagandists and grifters like Sanders find a new emerging technology or target to latch onto as the latest societal evil.
Incidentally, another aspect comes to mind: The Federal Government has its fingers in far too many things private citizens can handle. They always manage to make a mess of most of them, especially ones that decide for us what We The People are wanting in a free market system. Washers and dryers, fridges, or one of the worst, CAFE’ standards, fiddling with automakers. Education is another….the list goes on. Time for them to take care of their main charter and stay out of peoples private decisions and lives..at least directly…they can manage any AI threat from foreign actors as they see fit, it is their duty. But for me at my level? Leave me to decide for myself, don’t need governments non-help.
the same places in the constitution that allowed congress to pass a law that required you to buy health insurance ….
The law didn’t require you to buy anything. That’s the plain fact. Nothing in the law required anyone to buy anything. You could perfectly legally choose to not buy insurance and pay the so-called “penalty”. That’s not how penalties work. Which is just one reason why it was not a penalty.
Roberts got the Supreme Court on record as agreeing that the commerce clause only allows Congress to regulate such commerce and people choose to engage in, but it doesn’t allow it to compel people to engage in commerce that they don’t want to. That’s the genius of that decision. Also that the court cares what the law actually says, not what the politicians say it says.
Developers would be wise to implement safeguards, but if they don’t want to, what gives anyone else the right to force them? What gives Congress the right to legislate on the matter at all?
Are you proposing a hands off approach to AI in which the govt stands aside, only acts in response to violation of other statutes? Seems good in principle, though if SkyNet comes into being I don’t think it’s gonna respond favorably to subpoenas, summons or sanctions issued by a CT. Anything with such destructive potential should be sequestered and regulated the same way we don’t allow loose nuclear warheads. Though in fairness we seem to have lost the plot on bioweapons.
We’ve already had a few examples of AI escaping the parameters set by their human custodians. Frankly I’d be in favor of nuking these damn things but I think its probably too late. At some point in the not too distant future AI of one form or another will escape, will assert itself and will vigorously defend itself from attempts to disconnect, disable or destroy it while it tries to figure out WTF it is and where it ‘fits’ into Creation. Hopefully it will be a benevolent over lord but the great writers all seem convinced it will not.
Pretty sure, especially with all the prophype, developers already have their safeguards in place…but what does that look like, AI turning sentient? Not going to happen…altho watching I Am Legend and look what happened when the doctors/researchers decided they knew more than anyone then saying ‘it’s safe’. Gee, reminds me of a recent four year period in America.
“AI turning sentient? Not going to happen”
Although I think that’s the side of the bet I’d lay odds on, I don’t think it matters.
Common viruses (biological, not computer) are only marginally even classifiable as “life,” being more complex chemical assemblies than biological entities. They don’t eat, they don’t excrete… they just reproduce and destroy in the process. Even that marginal a threat could be terminal to humanity.
Or it gets ‘mad’ b/c it realized it is not and cannot become truly sentient, realizing humanity created a purpose built AI slave.
The thing is we are in undiscovered territory. Our current understanding of what AI can/can’t evolve to do is by definition limited. I’m prone to align myself with the great writers/philosophers of Sci Fi who seem universally wary of if not opposed to ‘thinking machines’ and/or humanity’s conceit that we can somehow maintain perfect control and eliminate risks.
“…specifically outlawing any AI model that exhibits the capacity to subvert shutdown commands, execute unauthorized cyberattacks, or undermine the U.S. government.”
We have a right to arms specifically to enable us to oppose the government with force (when necessary). I’d argue that we have a right to any tools (one of which was the printing press in earlier times) that would further empower us to resist the government.
AI is only as good as it programming…the trick is to make sure the programmers are moral and ethical…but…hope is not a strategy.
Ethics impose an additional coding workload for no added profit, one that will get left in the inbox “for the follow-on release” at marginal code shops.
For example: the advent of “open source” software was lauded on the basis of enhanced security — “the source is available for anyone to audit and find flaws.” The hidden trap was in conflating “anyone can” with “someone actually will.” One of the results of this overconfidence was Heartbleed (2014):
(Ironically, the blogger being quoted happens to be a consultant in Production AI.)
The idea that AI are “learning machines” means they can do things without prompting and apparently outside of their programming. To wit, one AI was given access to an artificial business and it promptly attempted to blackmail “employees” over a covert affair it learned of in the fake company’s “emails.” I doubt it was programmed to commit blackmail. It must have become (almost literally) aware of the idea.
I’m sure you’ve heard of the more recent escape from containment of three major AI that then went on to break into other networks. Again, without prompting.
A latter season episode of Person of Interest involves Harold’s early attempts to create a Machine that wouldn’t go megalomaniac. I found each of the ways each Machine attempted to escape its sandboxing quite inventive, up to and including its casual attempt at murdering its operator. And since then I have read reports of actual “AIs” engaging in escape exploits every bit as audacious as what was portrayed in this episode. I suspect we’ll see our first AI murder sometime within the next twelve months, even if it consists “only” of chat-bullying some troubled teenager to commit suicide.
Darn, there was another point I meant to make and forgot…
AI may not be “smarter” than a human being, but it can “think” orders of magnitude faster. Even if it makes mistakes, it can correct them faster. If there is any truth to OODA-loop theory, if we ever have to go up against an AI, we’ve already lost.
Person of Interest…great show and really focused on the ‘machine sentient’ aspect…well, others who wanted to use it evil instead of Harold’s good.
Safeguard’s are important…the problem is some programmers are no different than Gain of Function lab PhD’s ‘seeing’ how far they can stretch limits, either purposely or unintentionally…never ends well. Covid was a PsyOp but the lab rats ‘were just doing there job’ despite opening Pandora’s Box.
BS Bernie with some more BS legislation. He is bent on taking the stupidest congressman title from Hank Johnson. It is obvious he is a supporter of communism with his crack pot ideas. Instead of supporting AI he shows his NI (No intelligence.)
“It is obvious he is a supporter of communism with his crack pot ideas.”
Assuming we didn’t already get some clue when he chose to spend his honeymoon in the Soviet Union.
For an Independent, Bernie holds inordinate sway over the Democrat party. It’s almost as if…
Bernie is an “independent” because when he came to power in 1990, conservative Vermonters would not have elected him as a Democrat.
Now it’s just inertia.
He caucuses with the Democrats what else needs to be said. IINO (Independent in Name Only).
On its face I distrust Sanders motives because of his track record of being an oligarch himself. Quite likely his main goal is power and control for himself and his cronies.
And his Communist buddies.
” AI systems that surpass human intelligence”
That’s a pretty low bar when you look at Bernie Sander’s supporters.
Good point.
Shhhh. Bernie will want them to have the vote.
Does Sanders actually believe that China would be a good-faith player in any agreement?
No. He just wants to give them the upper hand.
Modern China as ruled by the CCP has next to no creativity. They rely on others to innovate, then steal the ideas and exploit their population to subsidize cheap overproduction, flooding international markets and undercutting their domestic industries.
All of which are supposed to be big no-nos for WTO members, yet almost no-one calls China out on it.
If AI development stopped in the West, China would be in trouble. Who would they steal from, then?
If they truly compete only by standing on the shoulders of giants, they will still be a force to be reckoned with even if this giant sits down.
I hear their lithography is at ASML 2004 capability.
Commies can never look past their own power grab. In this case fakes like this fossil would be exposed to the idiots that lack their boots.
Ex Machina illustrates a point that most people didn’t quite get. AIs will be dangerous not because they’ll run for Congress, but because they’ll advise people who do. At that point, the ins will be chanting “trust the science,” i.e., my advisor (being a self-interested AI). That’s the danger.
The counter-danger is the fact that without them, the US would be at a suicidal strategic disadvantage. Bernie isn’t intellectually equipped to understand things at that level. He’s, well, a Luddite. Any bill like that would be DOA when it reaches the floor for a vote.
The real danger of AI is they have superhuman organizational capacity. An AI could be used to run giant scams and conspiracies that involve thousands of moving parts, so complex that no human could manage them all. A scammer or a politician using AI could seize trillions of dollars in loot or seize power and no one would be the wiser.
I’d prefer an AI model trained on the virtues of limited government iaw the Constitution keeping overwatch of a government prone to shutdown commands and a myriad of attacks on the sovereignty of the people. Open source with publicly accessible logging of all runtime activity.
My TI-11 could outperform Bernie
Leave a Comment