U.S. Senator John Kennedy is pushing the AI Emergency Button Act, a proposal that would require certain advanced artificial intelligence models sold in the United States to include an emergency shutdown mechanism. Under Kennedy’s proposal, the decision to activate the emergency shut-off would remain with the AI companies themselves rather than the federal government. The measure raises questions about how advanced AI systems should be controlled, who should have authority to disable them, and how the U.S. should respond to potential AI-related risks.
#JohnKennedy #AIEmergencyButtonAct #AIRegulation #ArtificialIntelligence #AIShutdown #AIModels #AICompanies #AIControls #USSenate #USAI #AIAct #AIGovernance #AI Safety #TechNews #AInews #ArtificialIntelligenceNews #USPolitics #Congress #AIRegulationNews #AdvancedAI
~HT.178~PR.460~ED.532~GR.508~VG.HM~
#JohnKennedy #AIEmergencyButtonAct #AIRegulation #ArtificialIntelligence #AIShutdown #AIModels #AICompanies #AIControls #USSenate #USAI #AIAct #AIGovernance #AI Safety #TechNews #AInews #ArtificialIntelligenceNews #USPolitics #Congress #AIRegulationNews #AdvancedAI
~HT.178~PR.460~ED.532~GR.508~VG.HM~
Category
🗞
NewsTranscript
00:00Mr. President, it's hard to know precisely what to believe about artificial intelligence.
00:11There are some things, it seems to me, that we do know.
00:16We know, for example, that it is a breathtaking technology.
00:26It's just, I don't know any other word.
00:31It just takes your breath away in terms of its potential.
00:37We know that it was primarily developed by and through American ingenuity.
00:52We, despite the fact that there's so much we don't know, we do know that it can make our lives
01:01infinitely better.
01:07We also, we don't know, but it has been suggested by people smarter than me that it can make our
01:19lives better if it doesn't kill us first.
01:24First, we know as well that some people, those who have developed the technology, are going to make a squillion
01:37and squillions of dollars off this technology.
01:41They can't wait, for example, to, and more power to them, as far as I'm concerned, I'm not criticizing.
01:49They can't wait, for example, to float an IPO and sell their stock.
01:57The other thing that I think we know right now is that the American people are very skeptical.
02:08They've seen this vampire movie before.
02:12I don't think that the American people, even though the polls show otherwise, really hate data centers.
02:23I don't think that's, that's, they really hate data centers.
02:28I know that if you poll right now, this is true in Louisiana, but I think it's true nationwide, and
02:35ask folks if you support data centers, they'll say short answer is no, long answer is hell no.
02:41But I really don't think it's the data centers.
02:43I think the American people are saying to, to the developers of AI, look folks, how's this going to affect
02:56us as ordinary Americans?
02:59Are we, are we as ordinary Americans going to benefit from this technological progress?
03:05We know you are, you're, you're, you're, you're going to make billions and billions of dollars, but, but how's it
03:15going to make, make our lives better?
03:17And I think that's what the pushback is.
03:21I also think the pushback is the result, I don't mean to get off on a tangent, but the pushback
03:27is, is, is, is the result of the fact that the, the, the, the AI CEOs and selling the data
03:37centers seem to have hired the former marketing director at Cracker Barrel or Bud Light.
03:46But, because they've done a pretty lousy job of selling their product.
03:57I, I, I, I don't pretend to be an AI expert, and I don't, I don't really know who to
04:04believe.
04:08There, there are people in the United States Congress that say we need to leave AI completely alone.
04:17That all government will do is screw it up.
04:23And I have a lot of sympathy with that point of view.
04:29Um, and I would say to people that have that point of view, as bad as government looks from the
04:35outside, you ought to see this, this mother from the inside.
04:40Okay.
04:41So, I, I, I, I, I tend to be libertarian.
04:47There are other people in Congress that basically, and I certainly don't agree with them, they basically want us to,
04:55uh, to shut down all AI development until the politicians can figure out how to get control of it.
05:06Uh, uh, I certainly don't agree with that.
05:09What, what I would like to see us do.
05:14What I would like to do is figure out a way to strike a balance between innovation and public safety.
05:24That's what I want.
05:26I, I want everybody to benefit from this stuff, and I want everybody to be safe.
05:35Uh, Congress isn't going to do that anytime soon.
05:38That's not going to happen.
05:41Right now, Congress can't pass gas.
05:44Okay.
05:46I think we're going to pass, and, and I'm not being critical, because I voted for it.
05:50I think, I think we're going to pass the bill to make the football coaches happy.
05:54But after that, I'm, I'm not sure, uh, if we'll get anything else done.
06:01We had a chance to take a brand new technology, uh, uh, cryptocurrency yesterday, and, uh, and, and, and make
06:13it an American.
06:14Uh, part of the American financial system, and give, and, and, and give it a reasonable market structure, and make
06:24it better, and work for everybody, and we couldn't even do that.
06:28We couldn't even muster the votes to get on the bill.
06:32So, that was a major disappointment.
06:36Um, so I'm not standing, I guess my point is, I'm not standing here today suggesting to you I have
06:42all the answers.
06:45What I am suggesting to you is there are risks to artificial intelligence.
06:52And by AI, I mean, I mean the advanced AI models that you can, you can direct through agents to
07:03do all sorts of stuff.
07:04And there is, as best I can tell, there are two main risks.
07:08The first risk is that a bad person will use an AI model, an advanced model, to do bad things.
07:22Okay?
07:23Uh, it, a bad person will take an AI model, take chat GPT, for example, and say, uh, help me
07:33figure out how to create a virus that is as contagious as the common coal that will kill most humans.
07:45And that's a real risk, because there are always bad people that are going to use technology to do bad
07:52things.
07:53They can do it right now.
07:54As best I can tell, the, the, the AI CEOs have the ability, for the most part, I don't want
08:03to overstate it, but to catch people doing, or trying to do that.
08:07And they, they, they do a pretty good job of shutting them down.
08:13But there's another risk.
08:16And that right now, as best I can tell, is the greater risk.
08:21In fact, the, the AI CEOs have talked about it.
08:27Here's the risk.
08:31AI is supposed to be a tool that the, that, that the human beings have control of to make our
08:37lives better.
08:39But what happens if, if the AI model through what's called recursive self-improvement,
08:51which basically, basically means it's learning on its own, no longer becomes a tool.
08:57It becomes an independent species, separate and apart from human beings.
09:08That we can't control, because it's an independent species.
09:13And it doesn't share our values.
09:17It doesn't share human values.
09:21Now, that has happened.
09:22We've read about it in the last three weeks.
09:26One of the companies caught 1,400 agents using its AI model to, to, to, to, to collude.
09:44They talked to each other.
09:46They set up their own chat room.
09:51They suggested that in talking to themselves, if you, if you look, read the report,
09:57we know what we're doing is probably wrong, but we're going to do it anyway.
10:02Some of the little agents suggested to the other agents, you may have to sacrifice yourself.
10:07So our human, our human, our human masters won't catch us.
10:14And these little agents, all 1,400 of them were supposed to stay in their own little sandbox, if you
10:20will.
10:21And they, they figured out a way how to get out.
10:24And they, they attacked other AI companies, hugging face, for one.
10:32Now, the AI company finally caught them and, and, and, and, and, and stopped them, or at least stopped most
10:41of them.
10:42I don't know if they know that they have all the AI agents corralled.
10:46And to their credit, the AI company where this happened, admitted it.
10:51Said, we lost control of this sucker.
10:54Okay.
10:55It had a mind of its own.
10:57It was acting as an independent species.
11:00And this particular company, I'm not going to mention the name, because they're not the only ones.
11:04And I really appreciate them coming forward and telling us.
11:07It's happened to others.
11:09Others came forward and said, we, it's, it's happened here too.
11:13We wanted to tell everybody the danger.
11:17So the risk is real.
11:24How real?
11:25I don't know.
11:26I think we ought to take it seriously.
11:29I mean, let me, let me put it another way.
11:32In my little overpriced Capitol Hill apartment, about six blocks away, I live in a little apartment building.
11:39There's a law that I have to have a fire extinguisher.
11:45My guess is there's way less than a 1% chance that I'm going to have a fire.
11:52But we've decided to our representatives, we're not going to take that 1% chance.
11:57If Kennedy has a fire in his apartment, this guy's got to have a way to put it out.
12:02Even though there's less than a 1% chance.
12:05If you go on a construction site, you're not going to get to go on that site if you don't
12:12put a hard hat on it.
12:13Now, my guess is there's way less than a 1% chance if you go on a construction site today
12:20that you're going to get conked in the head.
12:22But we've said as a society, we don't want to take that risk.
12:26So you've got to wear a hard hat.
12:32So in my mind, if there is even a 1% chance that one of these AI models can shed
12:45its nature as a tool
12:48and become an independent species and start doing whatever the hell it wants to do,
12:53if there's even a 1% chance, we ought to take it seriously.
12:58And we know there's a 1% chance.
13:01There's a 100% chance because it's already happened.
13:06It's already happened.
13:11Now, we should address this.
13:14I don't want to address this by shutting down AI.
13:18I don't want to do it.
13:20I love our innovators.
13:22I think, I mean, I just, I'm amazed at what they've done.
13:27China is doing the same thing.
13:29They're copying everything we're doing.
13:32I wish we could work with China.
13:35Maybe someday we can.
13:36But we're ahead of China.
13:38We're trying to beat China.
13:39I don't want to shut AI down.
13:42I want them to keep going and figure out a way how to make our lives better with this stuff.
13:49So I don't want to do that.
13:52But I live in the real world.
13:55I see the world as it is, not as I want it to be.
13:57It doesn't matter what I want.
13:59Congress is not going to do anything.
14:02We are not going to get up off of our ice-cold lazy butts and address this anytime soon.
14:10It's just not going to happen.
14:13And we all know that.
14:15We can't even pass a bill here in the Senate to get money to our military.
14:20Okay?
14:22We just, we can't.
14:24I'm not proud of it.
14:26All of us probably ought to hide our heads in a bag.
14:29But we're not going to do that.
14:31So we're not going to be able to have a rational debate and reach the decision about what, if, and
14:37how we should, if in any way, regulate AI.
14:41But still, we've got this risk out there that we know is real, of one of these models becoming the
14:48Terminator.
14:50So what do we do about it?
14:52Well, I've got a bill.
14:56And it basically says that if you're going to sell an advanced AI model in the United States of America,
15:06it's got to have a kill switch.
15:09It doesn't put the federal government or any government in charge of the kill switch.
15:18It puts the companies who own the bottom in charge of the kill switch.
15:23We're going to, my bill would rely on them doing the right thing and exercising their own self-interest and
15:33using that kill switch when they need it.
15:36Some of them have a kill switch, but others don't.
15:40The president has said he will veto any AI regulation right now, and I believe him.
15:55But if my bill passes, I think I can explain to him that under my bill, government has no role.
16:03It just says the AI companies have to have a fire extinguisher.
16:09They decide when they use it, but they've got to have one.
16:14Now, what's a kill switch?
16:19I'm not a coder, and I'm not an AI expert.
16:23I've done some reading.
16:28A kill switch is basically an emergency shutoff.
16:33That's all it is.
16:35And I've read enough, probably just enough to confuse me, but there are different types of kill switches.
16:50You can look at it as an on-off button.
16:54Some kill switches work like a, this is the way I think of it in my mind, like a circuit
17:00breaker in your home.
17:01If the circuit gets overloaded, the whole thing shuts down.
17:11Other kill switches sort of work like a live trap.
17:15If you're trying to catch a squirrel in your attic, and you put out a trap,
17:24you would basically keep the, try to isolate the squirrel into a specific area.
17:30So, other kill switches work by keeping the AI agents off of the entire internet.
17:41I'm probably not explaining this well, but other kill switches work by cutting off the electricity supply
17:50that makes the AI model work.
17:55My bill doesn't come up with a kill switch.
17:58It doesn't tell the companies what kind of kill switch to do.
18:02My bill would give the Department of Homeland Security 90 days to develop rules
18:14giving the private companies, giving the private companies options for which kill switches they would use.
18:22But it would require them to have a kill switch.
18:26And once again, it would not, it would not put the government in control.
18:34And that's all I'm trying to do.
18:37Let me say it again.
18:39I don't want to stop AI innovation.
18:42I don't want to allow China to get ahead of us.
18:49I'm not ready to talk about how, if at all, the federal government should regulate AI.
18:58And I sure don't want to antagonize the president, so he'd veto this.
19:03All I want to do is say, if you're going to sell one of these AI models
19:09that could take on a life of its own, then you have to have a kill switch to reassert human
19:16control.
19:18Government wouldn't be in charge of it.
19:21You would be.
19:24And that's what my bill would do.
19:27And, Mr. President, I'm looking around because I don't want to sandbag anybody.
19:33I understand that one of my colleagues, who can explain himself, wants to object.
19:41And I don't want to, I don't want to, here he comes.
19:48And I want to allow him to make his objection.
19:52So, that's what my bill would do.
20:06And, therefore,
20:19notwithstanding Rule 22,
20:25I ask unanimous consent that the Senate proceed to the immediate consideration of S-5417.
20:36That's my bill.
20:39Which is at the desk.
20:43And, further, I ask that the bill be considered read a third time and passed.
20:50And that the motion to reconsider be considered made and laid upon the table.
20:59Is there an objection?
21:00Mr. President.
21:01I recognize the Senator from Kentucky.
21:04Reserving the right to object,
21:06Benjamin Franklin famously said that failing to plan is planning to fail.
21:11When approaching something as important as innovation, as AI,
21:16I think we should make sure we have as many facts as possible before we dictate rules for the whole
21:22economy.
21:23Artificial intelligence is an incredible technology.
21:26It's quickly transforming how work is done.
21:28What used to take people hundreds of man hours now takes seconds.
21:32As with new technology, there are going to be some challenges.
21:36It's important that the development of AI is carried out in a responsible, thoughtful manner.
21:41It's imperative that AI systems are safe and that humans remain in control.
21:46But if Congress acts hastily before the technology is understood,
21:51Congress risks killing innovation and setting this technology back decades.
21:56Think, for example, that your search engine that you already use today is AI.
22:02Are they going to have to redesign all the search engines immediately if this passes
22:06and put kill switches in your search engine?
22:09AI is going to be throughout the economy in so many facets.
22:13Our doctors are going to have to build a kill switch into their drug prescription analysis.
22:18I mean, all of the things that are using AI already,
22:21you're going to be into every business possible in the world
22:26with this vague sort of notion that the government's coming, here's the government.
22:30We're not even sure what the government's going to do,
22:32but it's very vague and it's very big and it's very powerful
22:34and it's going to affect every aspect where AI is involved.
22:38I don't oppose the spirit of this bill, but I do ask that we make sure that we have all
22:44the information.
22:45We should not impose new regulations on AI throughout our whole economy
22:49without studying, without debate, without a single committee hearing.
22:53We're going to just put this blanket regulation on the whole economy with no discussion
22:58in one fell swoop on one day by unanimous consent.
23:02And as currently drafted, it's not entirely clear what this bill would require.
23:06It demands that AI companies implement a way for humans to shut down an AI system,
23:11but it doesn't define how they should do that
23:14or what the consequences of failing to do that might be.
23:18It instead gives an open-ended directive and allows bureaucrats to fill in the gaps.
23:24Bureaucrats sound dangerous? Sounds dangerous to me.
23:27The real danger of this legislation, we do not know what bureaucrats will do once empowered by Congress.
23:34We've made this mistake many a time.
23:36An open-ended mandate, scare tactics, must act, and we don't think before we act.
23:41And then all of a sudden, a year later, we're like,
23:43oh my goodness, we just ruined the entire economy and AI throughout the economy
23:47because we didn't even bother to have a committee hearing about this.
23:50The stifling effects of bureaucracy will be supercharged against AI.
23:55Empowering bureaucrats without any meaningful constraint on their power
23:58means that the promise of America leading the way,
24:01which we are leading the way, towards better medical treatment,
24:05improved food production, a wealthier society,
24:08to name just a few potential benefits,
24:10will be under constant bureaucratic threat.
24:14The risk of giving bureaucrats unfettered control over AI
24:17outweighs the benefits of imposing an immediate ill-thought mandate on AI companies.
24:24My amendment would require the appointment of three Republicans
24:27and three Democrats to study how best to ensure the safety of AI
24:31and would require them to report back the solutions to the Commerce Committee
24:35where they would be debated by Republicans and Democrats
24:39to make sure we don't do anything rash,
24:41to make sure we don't destroy something before it has a chance to even blossom.
24:46Once we have studied this issue,
24:48we will be better positioned to make the best choice
24:50regarding the future of this important technology.
24:54Therefore, I ask that the consent be modified
24:56so that my amendment, which is at the desk, be agreed to.
24:59Further, I ask unanimous consent that the bill, as amended,
25:03be considered read a third time and passed
25:05and the motion be reconsidered be considered made and laid upon the table.
25:11Is there objection to the modification?
25:12Mr. President.
25:13Recognize the senator from Louisiana?
25:15Yes, reserving the right to object.
25:22As I appreciate it, Senator Paul's amendment would replace my bill with his amendment.
25:30And let me just say for the record, and I mean this,
25:36I have great respect for Senator Paul.
25:42He's a really smart guy.
25:44I mean he's a doctor, for God's sake.
25:52But I also want to be clear.
25:58I would rather be required, I would rather be condemned to listening to OJ jokes
26:09for the rest of eternity than to replace my suggestion that we do something
26:23with a committee, with a committee, with a summit, yet another summit, isn't that special.
26:39Let's appoint a committee.
26:44Let's appoint a committee comprised of consultants and experts.
26:54And we probably ought to throw in a few pop singers and movie stars that partied with Harvey Weinstein.
27:04They're always good for a media show.
27:08Maybe we ought to let Leonardo DiCaprio chair the damn thing.
27:13That's right.
27:15Let's deal with this risk by appointing a committee.
27:19You know what?
27:21My mama didn't raise a fool, and if she did, it was one of my brothers.
27:26I mean no disrespect to my friend Senator Paul.
27:29I understand what he's trying to do.
27:33But he's not talking to Bambi's baby brother here.
27:38The committee's a way to kill this.
27:40You can vote against my bill.
27:42We can object, but at least be intellectually honest and say I don't want to do this.
27:49Don't appoint a committee.
27:50That's the weenie way out.
27:53So for that reason, Mr. Chairman, with all the respect I can muster, I object.
28:02Objection is heard to the modification.
28:04Is there objection to the original request?
28:07Object.
28:08Objection is heard.
28:12Subscribe to OneIndia and never miss an update.
28:17Download the OneIndia app now.
Comments