Latest / Good Morning BT with Bo Thompson & Beth Troutman / Theresa Payton: Thoughts on A.I. Industry "Self-Regulation"
Transcript
- 0:00Speaker 1: Good morning.
- 0:00Speaker 2: Good morning.
- 0:01Speaker 3: I listen to you guys every morning, and I travel
- 0:03Speaker 3: all over the U.S.
- 0:04Speaker 2: You guys keep it in the.
- 0:05Speaker 3: Middle of the road, as my dad used to tell me.
- 0:07Speaker 2: This is Good Morning BT.
- 0:10Speaker 3: And judging by the smile, it's real.
- 0:34Speaker 2: Well, not he, but she.
- 0:36Speaker 4: And I say this because our next guest has been
- 0:38Speaker 4: in Austin and Waco, Dallas this week.
- 0:42Speaker 2: I mean, she is making the trek through Texas.
- 0:44Speaker 4: Our longtime cybersecurity expert, Teresa Payton, joins us this morning.
- 0:48Speaker 2: Good morning to you.
- 0:50Speaker 3: Good morning. I mean, I do love Texas. It is
- 0:53Speaker 3: a really lovely state. I mean, North Carolina, of course,
- 0:58Speaker 3: is my favorite, but I do really love Texas. People
- 1:01Speaker 3: here are so great, so wonderful. In Waco today, on
- 1:05Speaker 3: my way to Richardson tonight, and it's been.
- 1:07Speaker 2: Quite a week.
- 1:08Speaker 4: Well, you said it quite a week, and yesterday in particular.
- 1:11Speaker 4: This is after the meeting at the White House with
- 1:14Speaker 4: President Trump and all the tech heads about AI or
- 1:19Speaker 4: SI or whatever we're calling it these days. But before
- 1:22Speaker 4: we get your take on what happened, here is President Trump. Mr.
- 1:25Speaker 5: President, did anything that you heard today, did anything change
- 1:28Speaker 5: your mind, given the fact that you said you don't
- 1:31Speaker 5: want to see guardrails, you don't want to see AI
- 1:33Speaker 5: slowed down? Anything that you heard today change your mind
- 1:35Speaker 5: on that?
- 1:35Speaker 1: Well, I think I'm seeing tremendous self-policing. And they understand
- 1:39Speaker 1: that they have to self-police. And it's very hard for
- 1:41Speaker 1: somebody to come out and say, get into these models.
- 1:44Speaker 1: These models are very complex. And again, I believe they're
- 1:47Speaker 1: going to be used for the good.
- 1:48Speaker 2: And when they're not, we're going to be able to
- 1:49Speaker 2: nab them.
- 1:50Speaker 1: But there's going to be a tremendous self-policing aspect. We're
- 1:54Speaker 1: also thinking about forming a committee of sorts, where we
- 1:58Speaker 1: put maybe 10 people on that committee. It could be
- 2:00Speaker 1: from that group, so the committee can watch over The
- 2:05Speaker 1: whole enterprise.
- 2:08Speaker 5: You've also indicated that Democrats are talking about concerns about
- 2:11Speaker 5: AI as a hoax. Do you believe that after meeting
- 2:14Speaker 5: with all these AI tech leaders?
- 2:15Speaker 1: Yeah, I do. I think they would love to see
- 2:16Speaker 1: it shut down, just like it was a hoax the
- 2:19Speaker 1: 12 years for global warming, which they changed right now
- 2:22Speaker 1: to climate change.
- 2:23Speaker 5: Even though Bill Gates says that AI could kill up
- 2:26Speaker 5: to a billion people?
- 2:26Speaker 1: Well, now he's not saying that anymore.
- 2:28Speaker 5: He said that this Sunday.
- 2:29Speaker 1: Well, I don't care what he said Sunday, because I
- 2:31Speaker 1: can tell you yesterday he didn't say that at all.
- 2:33Speaker 5: Mr.
- 2:33Speaker 2: President!
- 2:36Speaker 1: He actually went the opposite. Look, it's one voice. He
- 2:39Speaker 1: actually went the opposite. No, we're looking at this for good.
- 2:43Speaker 1: It's also maybe the biggest industry. They say bigger than
- 2:47Speaker 1: the Industrial Revolution.
- 2:48Speaker 4: Think of it.
- 2:49Speaker 2: You know what?
- 2:50Speaker 1: We're leading by a lot, and we're going to keep
- 2:52Speaker 1: it that way.
- 2:52Speaker 4: All right, that's the president with some questions after this
- 2:55Speaker 4: group got together yesterday.
- 2:57Speaker 2: Teresa, what do you make of what you heard and
- 3:00Speaker 2: saw yesterday?
- 3:01Speaker 3: Well, that was a lively press conference, that's for sure. Okay,
- 3:06Speaker 3: a lot of things. And this was on the minds
- 3:08Speaker 3: of everybody at Baylor University last night. So a couple
- 3:12Speaker 3: of things. One, I did finally get to see the document.
- 3:15Speaker 3: So there is a one-page document that is the White
- 3:17Speaker 3: House Accord on Superintelligence. And it is voluntary, but they
- 3:23Speaker 3: all got into a room, signed the document. I think
- 3:26Speaker 3: it's interesting.
- 3:27Speaker 1: There's only.
- 3:28Speaker 3: six signatures plus the president. There are a lot more
- 3:31Speaker 3: people involved in AI than the six people who signed.
- 3:36Speaker 3: So that to me, I'd like to see another signature
- 3:38Speaker 3: page be added today. But this is a step in
- 3:42Speaker 3: the right direction. I'd like to see more. I'd like
- 3:45Speaker 3: to hear more from the AI companies. There's four kind
- 3:48Speaker 3: of key components to the White House Accord on Superintelligence,
- 3:53Speaker 3: One is robust internal controls. Well, hello, you should have
- 3:56Speaker 3: had those to begin with. Another one is empowering an
- 3:58Speaker 3: internal team to ensure all the controls and monitoring are
- 4:02Speaker 3: operating as intended. The third is they agree to have
- 4:05Speaker 3: an independent external auditing and evaluators carry out assessments. And
- 4:10Speaker 3: the fourth is they're supposed to designate an independent committee
- 4:14Speaker 3: of the board of directors to oversee and receive reports.
- 4:18Speaker 3: These are all best practices. Most companies do anyhow, but
- 4:21Speaker 3: it's great to see they all got together, agreed to
- 4:24Speaker 3: these four things, and they signed their names to it
- 4:27Speaker 3: along with President Trump. However, there's still more to be done,
- 4:31Speaker 3: but this is definitely a step in the right direction.
- 4:33Speaker 6: Are you concerned at all about the idea of the
- 4:37Speaker 6: tech companies self-regulating? I mean, we've seen in the past
- 4:41Speaker 6: this go horribly wrong. I mean, the first thing that
- 4:44Speaker 6: comes to my mind is, the 2008 financial crisis, or
- 4:48Speaker 6: maybe the Volkswagen crisis that happened in 2015 with the
- 4:52Speaker 6: emissions scandal. Are you concerned that there's a conflict of
- 4:56Speaker 6: interest here of profit versus public safety versus publicly traded companies? Like,
- 5:04Speaker 6: Is there a problem here? And do you see these
- 5:07Speaker 6: folks actually being capable of self-regulating? Because we're talking about
- 5:11Speaker 6: the same people that created social media and allowed misinformation
- 5:16Speaker 6: to impact elections around the globe. I mean, we're talking
- 5:20Speaker 6: about people who have made some questionable choices.
- 5:24Speaker 3: Yeah, I mean, the one thing that strikes me is
- 5:28Speaker 3: that many of the individuals that signed this document did
- 5:33Speaker 3: not spend their career working in heavily regulated industries. I
- 5:38Speaker 3: worked in banking. And in banking, you have kind of
- 5:42Speaker 3: the banks will say, these are all the things we're
- 5:44Speaker 3: doing because it's just the right thing to do. And
- 5:47Speaker 3: then you also have the regulators who say, and these
- 5:49Speaker 3: are the additional things you will be doing because we're
- 5:51Speaker 3: going to tell you, you have to do it. And
- 5:54Speaker 3: so it occurs to me, you're right, Beth, like the
- 5:57Speaker 3: people that signed this document, I don't believe many of
- 6:00Speaker 3: them spent 20 plus years In a heavily regulated industry,
- 6:04Speaker 3: social media is not heavily regulated. There are regulations, but
- 6:08Speaker 3: not as many. Checks and balances and regulators and things
- 6:12Speaker 3: that they have to follow. They don't have to do,
- 6:14Speaker 3: you know, file different types of reports other than being
- 6:16Speaker 3: a public company and doing things for the SEC. But
- 6:19Speaker 3: as it relates to safety and security, that has been,
- 6:22Speaker 3: you know, I think most Americans would say more needs
- 6:26Speaker 3: to be done on social media. Obviously, I do have concerns, Beth,
- 6:30Speaker 3: but at least them agreeing to this opens up the
- 6:33Speaker 3: door for more conversation to say, well, how are you
- 6:36Speaker 3: going to do that? And do you see this as
- 6:38Speaker 3: a competitive advantage that, you know, can you stake, you know,
- 6:42Speaker 3: a claim that you're going to be the safest and
- 6:44Speaker 3: the securest and the best at doing this? Because I
- 6:47Speaker 3: do believe consumers and businesses will follow. And then lastly,
- 6:51Speaker 3: I am concerned that that the Hill, if they do
- 6:55Speaker 3: indeed pass a law, it's going to be too late,
- 6:57Speaker 3: it's going to be too big and will be, and
- 7:00Speaker 3: it's not really going to be what the American people
- 7:02Speaker 3: and businesses are looking for.
- 7:04Speaker 4: Well, speaking of the Hill, Representative Ted Lieu yesterday put
- 7:08Speaker 4: a social media message out, says this, the White House
- 7:11Speaker 4: AI meeting is fatally flawed even before it begins. No
- 7:15Speaker 4: representatives from AI safety, states, working families, civil rights, rural
- 7:20Speaker 4: communities or parents who lost their children to depraved chatbots.
- 7:24Speaker 4: And so obviously this is going to be a bit
- 7:26Speaker 4: of a political issue anytime you have a gathering like yesterday.
- 7:29Speaker 2: What about that notion?
- 7:32Speaker 3: No, he's right that the voices weren't there. And it's interesting.
- 7:36Speaker 3: So I'm going to read you a different point of view,
- 7:38Speaker 3: but it actually makes the same point. That's the funny thing.
- 7:41Speaker 3: So David Sachs said only President Trump can convene all
- 7:44Speaker 3: the leaders of the top companies developing chips, data centers,
- 7:47Speaker 3: and frontier models for superintelligence. He goes on to say
- 7:50Speaker 3: other great things about it. I was honored to witness
- 7:53Speaker 3: history as the leaders of the Frontier Lab Company signed
- 7:55Speaker 3: the White House Accord on Superintelligence, accepting responsibility for the
- 7:59Speaker 3: safe development of their products, and imposing new internal controls
- 8:02Speaker 3: and external audits. And then David goes on to say
- 8:04Speaker 3: other things. And you're right. To me, when I looked
- 8:07Speaker 3: at the signature page and saw how short it was,
- 8:09Speaker 3: I thought, there's a lot of other constituencies and stakeholders
- 8:15Speaker 3: that are not represented by this. But maybe we can
- 8:18Speaker 3: hope there's a part two. And so maybe the next
- 8:21Speaker 3: step is let's hear from the people who should have
- 8:24Speaker 3: been on the second and third and fourth signature page.
- 8:28Speaker 3: I'll take this as a step in the right direction,
- 8:30Speaker 3: but I've got more questions than answers right now.
- 8:32Speaker 6: I know we're almost out of time here, but I
- 8:34Speaker 6: wanted to ask you really quickly, are we going to
- 8:36Speaker 6: start calling this superintelligence? Are we going to call this SI?
- 8:39Speaker 6: Because originally superintelligence was the hypothetical level of computer capability
- 8:44Speaker 6: surpassing human intellect. Are we all going to, as a world,
- 8:49Speaker 6: have to call it SI now?
- 8:52Speaker 3: Well, okay, so President Trump signed an order yesterday renaming
- 8:57Speaker 3: AI as superintelligence. I've got to dig into that a
- 9:02Speaker 3: little bit more, so I don't know. So by signing
- 9:05Speaker 3: an executive order, that does compel the departments and agencies
- 9:08Speaker 3: to do that. So the departments and agencies certainly will
- 9:12Speaker 3: start calling it superintelligence. The question is, you know, will
- 9:15Speaker 3: the commercial markets, will consumers... And will the rest of
- 9:18Speaker 3: the world say, sure, it's not AI anymore, it's SI?
- 9:22Speaker 3: I think the main thing that everybody should just be
- 9:25Speaker 3: thinking about is, as it relates to these automated algorithms, intelligence, data,
- 9:31Speaker 3: whatever we want to call these computer programs, tomorrow versus yesterday,
- 9:37Speaker 3: we do need governance and guardrails. And so, you know,
- 9:39Speaker 3: at least there's some acknowledgement in this letter that those
- 9:43Speaker 3: are needed. And then we'll just see what the kind
- 9:46Speaker 3: of the governance strategy is going to be about these agreements.
- 9:49Speaker 4: Is there an option for me to get SI, the
- 9:51Speaker 4: paper version, back in my mailbox once a week?
- 9:53Speaker 2: Yeah, and can we get the swimsuit edition?
- 9:55Speaker 3: So true. When you say SI, it's like Sports Illustrated. right?
- 10:00Speaker 2: That's what we're talking about here. That's what it will
- 10:02Speaker 2: always be to me. I don't care what anybody says.
- 10:04Speaker 4: Hey, we've got to stop it there, and we'll continue
- 10:06Speaker 4: to watch the fallout from what happened yesterday, and like
- 10:09Speaker 4: you say, maybe a part two coming. But thanks for
- 10:11Speaker 4: all your reaction and analysis.
- 10:13Speaker 3: Yeah, no, it's great to be with you, Beth and Bo.
- 10:15Speaker 3: Be safe out there.