Latest / Kvetch & Release / AI Whiplash: Pre-Crime, Fake Reporters, & Immortality or Human Extinction

Transcript

AI Whiplash: Pre-Crime, Fake Reporters, & Immortality or Human Extinction · Kvetch & Release

• 0:00
  Speaker 1: Right now, though, it's time for Lentke's Got a Take
• 0:03
  Speaker 1: with John Lentke. And good morning.
• 0:06
  Speaker 2: Good morning, Gene.
• 0:07
  Speaker 1: Well, I have that really incredible story in the latest
• 0:11
  Speaker 1: on the news. The guy who quit Anthropic, the AI company,
• 0:14
  Speaker 1: back on Tuesday, saying that artificial intelligence will be capable
• 0:19
  Speaker 1: of killing people by the end of the decade without
• 0:23
  Speaker 1: any human involvement. In other words, it'll just take over.
• 0:26
  Speaker 1: and do it all by itself. So you were talking
• 0:29
  Speaker 1: to me yesterday, and you wanted to talk about artificial intelligence,
• 0:33
  Speaker 1: the news and social media, and how it's kind of
• 0:37
  Speaker 1: taken over already. So when you read that a purported
• 0:41
  Speaker 1: news outlet literally created a fake journalist whose name I
• 0:46
  Speaker 1: can't pronounce, it looks like Lee Kwamedig, to pull stories
• 0:51
  Speaker 1: out of northern Canada, what went through your mind? I mean,
• 0:54
  Speaker 1: it's fake news.
• 0:56
  Speaker 2: Yeah, Gene, I mean, like, this is what's so crazy
• 0:58
  Speaker 2: about it, right? I mean, for years we know that
• 1:00
  Speaker 2: bad actors have used bot farms to be dumping trash
• 1:04
  Speaker 2: into our social media feeds, but now literally inventing synthetic
• 1:08
  Speaker 2: reporters with fake faces to scrape and steal real local news?
• 1:12
  Speaker 2: I mean, but we shouldn't be shocked about it. The
• 1:14
  Speaker 2: entire business model of generative AI is built on industrial-scale
• 1:19
  Speaker 2: IP theft. We know this through court cases that are ongoing.
• 1:23
  Speaker 2: They scrape human intelligence, creativity, and sweat. They strip the credit,
• 1:27
  Speaker 2: they repackage it, and they sell it back to us,
• 1:29
  Speaker 2: or they turn us into the product by providing it
• 1:32
  Speaker 2: for free and scraping our personal data to build more
• 1:35
  Speaker 2: of their algorithms. To pull that stunt, though, in northern Canada...
• 1:39
  Speaker 2: a region already targeted by foreign disinformation as a direct
• 1:42
  Speaker 2: hit on community trust. And the operators running these scams
• 1:45
  Speaker 2: don't care about journalism. They don't care about democracy. They
• 1:48
  Speaker 2: don't care about community. They care about free content being
• 1:51
  Speaker 2: used for clicks to get ad revenue. Consequences to democracies
• 1:55
  Speaker 2: and societies be damned.
• 1:57
  Speaker 1: And, you know, social media is robbing Canadian broadcasters blind.
• 2:03
  Speaker 1: Because a lot of the money is going into social
• 2:05
  Speaker 1: medias because they get clicks. And, you know, our local
• 2:09
  Speaker 1: Hamilton TV station canceled their 6 and 11 p.m. news,
• 2:13
  Speaker 1: not because they wanted to, but because the ad revenues
• 2:16
  Speaker 1: are going down. into these social media companies in the U.S.
• 2:20
  Speaker 1: They're literally, the U.S. social media people, they're literally robbing
• 2:24
  Speaker 1: us without a gun. So when I'm thinking about what
• 2:27
  Speaker 1: you just said, isn't this just the next logical step
• 2:31
  Speaker 1: in social media disinformation? Or is there something fundamentally even
• 2:35
  Speaker 1: worse about using synthetic identities to scrape news?
• 2:40
  Speaker 2: I think this is worse, Gene, because this is weaponizing
• 2:43
  Speaker 2: the identity to manufacture unearned trust. I mean, over the
• 2:50
  Speaker 2: past decade, we've mostly learned how to spot bots sharing
• 2:53
  Speaker 2: fake links or organized disinformation campaigns. But when you wrap
• 2:56
  Speaker 2: an AI scraper in a fake face and claim that
• 2:59
  Speaker 2: she's an Inuk reporter raised in Nunavut who understands the region,
• 3:03
  Speaker 2: I mean, that's cultural theft as well here, not just
• 3:05
  Speaker 2: IP theft. It's piggybacking on the credibility of real local
• 3:08
  Speaker 2: journalists who put in the work to cover the news
• 3:11
  Speaker 2: of their communities. So it completely poisons the well, and
• 3:14
  Speaker 2: if people in any community, let alone remote communities, can't
• 3:17
  Speaker 2: trust whether the person bylining a local story actually exists,
• 3:21
  Speaker 2: local news collapses.
• 3:22
  Speaker 1: Yeah. And local news is important. That's why we are
• 3:26
  Speaker 1: live and local in the region here on CKTV News.
• 3:30
  Speaker 1: You know, we've seen tech companies steal content, but what
• 3:34
  Speaker 1: about stealing our privacy? What do you make of Anthropix
• 3:36
  Speaker 1: building what they call a pre-crime tracking system for activists?
• 3:41
  Speaker 2: I mean, listen, Tom Cruise gave us Minority Report as
• 3:44
  Speaker 2: a dystopian warning, Gene. Silicon Valley treated it like a
• 3:48
  Speaker 2: pitch deck that they found on a, I don't know,
• 3:51
  Speaker 2: coffee table somewhere in a Starbucks in Silicon Valley. It's
• 3:54
  Speaker 2: no surprise that as AI embeds itself more into our society,
• 3:58
  Speaker 2: the national security state, wants to weaponize it. But once
• 4:02
  Speaker 2: these tech giants steal all of our data and IP
• 4:04
  Speaker 2: to build their model, the next step, as now we've
• 4:06
  Speaker 2: seen them discussing, is building predictive algorithms to monitor those
• 4:10
  Speaker 2: who push back on the state or power more broadly. So, yeah,
• 4:13
  Speaker 2: this is a corporate power grab. It's a state power
• 4:15
  Speaker 2: grab that will inevitably be used. And we're watching tech
• 4:19
  Speaker 2: giants attempt to or begin to turn public dissent into
• 4:23
  Speaker 2: quantifiable security threats before any crime has even been committed
• 4:26
  Speaker 2: in sending that information to the police. This is not
• 4:28
  Speaker 2: the same as a neighborhood watch.
• 4:30
  Speaker 1: Yeah, well, Anthropic frames itself as a safe, responsible, artificial,
• 4:35
  Speaker 1: intelligent giant. So how does a company justify building, like
• 4:39
  Speaker 1: you said, a minority report-style surveillance system?
• 4:43
  Speaker 2: They can justify it however they want to make themselves
• 4:45
  Speaker 2: sleep at night. I think they're sleeping at night on
• 4:47
  Speaker 2: incredibly nice mattresses made out of $ 100 bills. They're spending
• 4:52
  Speaker 2: millions on PR to tell us they're responsible, ethical guys
• 4:55
  Speaker 2: that are in the room. They're the good cops at AI.
• 4:57
  Speaker 2: But then they're turning around and hiring intelligence analysts to
• 5:00
  Speaker 2: build threat monitoring engines to target political dissent. And this
• 5:03
  Speaker 2: is classic cover. They say AI safety when they talk
• 5:06
  Speaker 2: to lawmakers to look good. But in practice, safety just
• 5:09
  Speaker 2: means keeping their corporate power safe. safe from public scrutiny.
• 5:12
  Speaker 2: It's terrifying in that sense. And it proves that no
• 5:15
  Speaker 2: matter how clean the branding is, these companies will always
• 5:17
  Speaker 2: choose surveillance over civil liberties. I mean, look, I mean,
• 5:20
  Speaker 2: one of the open AI started out as a not-for-profit,
• 5:22
  Speaker 2: and they very quickly went, you know what, maybe profit
• 5:25
  Speaker 2: is the way that we need to do things.
• 5:26
  Speaker 1: Yeah, that's so true. Well, if we look at the
• 5:29
  Speaker 1: big picture, on one hand, we're told that artificial intelligence
• 5:33
  Speaker 1: will do great things like stop biological aging. And on
• 5:37
  Speaker 1: the other side, experts are warning, just like this guy,
• 5:40
  Speaker 1: of a 10% chance without any human interaction whatsoever, it'll
• 5:44
  Speaker 1: start killing people. So what side of the hype are
• 5:47
  Speaker 1: you buying?
• 5:48
  Speaker 2: Someone made a joke to me on social media yesterday.
• 5:51
  Speaker 2: It is going to do both. It'll stop the biological aging,
• 5:54
  Speaker 2: and then it's going to kill us all. So we
• 5:56
  Speaker 2: get the best of both worlds. But no, I don't
• 5:58
  Speaker 2: buy either narrative at face value right now, because both
• 6:01
  Speaker 2: are being weaponized by people looking for their bottom lines.
• 6:04
  Speaker 2: Biotech firms claiming they can reverse or stop biological aging
• 6:07
  Speaker 2: are just looking to pump up their valuations and drive
• 6:10
  Speaker 2: stock prices. On the flip side, the AI apocalypse narrative
• 6:13
  Speaker 2: is a giant PR distraction right now, which we can
• 6:16
  Speaker 2: talk about because they don't want competition. So is the
• 6:19
  Speaker 2: technology making huge strides? Absolutely. It's solving impossible math problems.
• 6:24
  Speaker 2: I saw one last week that hadn't been solved in
• 6:26
  Speaker 2: like 90 years.
• 6:27
  Speaker 1: Yeah, yeah, yeah.
• 6:28
  Speaker 2: It's accelerating drug discovery. But we can't go letting shiny
• 6:31
  Speaker 2: promises or sci-fi horror stories blind us. And we don't
• 6:35
  Speaker 2: need to wait 10 years for a rogue AI to
• 6:37
  Speaker 2: destroy humanity. If we let AI destroy media integrity today
• 6:41
  Speaker 2: and we hand over control of our democracies to corporate
• 6:44
  Speaker 2: surveillance tomorrow, what's worth left to save anyway? So is
• 6:48
  Speaker 2: the risk real? Yeah. Are the benefits equally real? Yeah.
• 6:51
  Speaker 2: But I'm not going to fearmonger or hide under my desk.
• 6:54
  Speaker 2: We stopped doing that with nukes, and there's still tens
• 6:57
  Speaker 2: of thousands of them pointed at virtually every major population center.
• 7:00
  Speaker 2: But we don't let it ruin our day.
• 7:03
  Speaker 1: In the end here, is there something more at play
• 7:06
  Speaker 1: that we're missing, maybe?
• 7:07
  Speaker 2: I think so, Gene. I think what we're missing here
• 7:12
  Speaker 2: is that what we're really missing is that We're watching
• 7:17
  Speaker 2: textbook regulatory capture. Why are AI CEOs suddenly parading around
• 7:22
  Speaker 2: Washington and Ottawa and other global capitals crying that their
• 7:25
  Speaker 2: own technology, some in cases like weeks before their IPOs, hilarious,
• 7:29
  Speaker 2: is a nuclear-level threat to humanity? Well, because if they
• 7:33
  Speaker 2: can convince governments that AI is too dangerous to be
• 7:36
  Speaker 2: left open, the legislators will pass massive licensing hurdles that
• 7:40
  Speaker 2: pull up the ladder beneath them, keeping their competitors at bay.
• 7:43
  Speaker 2: And it could outlaw open source competition, kill off smaller startups,
• 7:48
  Speaker 2: and guarantee that a handful of megacorporations hold a permanent
• 7:51
  Speaker 2: monopoly over the most powerful technology on Earth. So they're
• 7:54
  Speaker 2: not trying to save humanity from AI today. They're trying
• 7:57
  Speaker 2: to save their market share from competition tomorrow.
• 8:00
  Speaker 1: Yeah. John, I miss the good old days of Facebook
• 8:03
  Speaker 1: when I would open my page and I could enjoy squirrels,
• 8:07
  Speaker 1: water skiing, and small boats.
• 8:10
  Speaker 2: Those were the days, Gene. Maybe we'll get back there,
• 8:13
  Speaker 2: but it doesn't seem like it.
• 8:15
  Speaker 1: Oh, geez. What a great take. Thanks, John.
• 8:18
  Speaker 2: Thank you, Gene.
• 8:19
  Speaker 1: Okay, Litke's got a take. He's back with us on
• 8:21
  Speaker 1: Friday morning at 9.15.