It doesn’t take a genius to manipulate the population. It just takes some mid-level AI chatbots with a mean streak.
That thought haunted Glenn Reynolds, the author of the new book “Seductive AI.” The tome doesn’t look to a near future in which artificial intelligence has a profound impact on our lives and culture.
‘The media actually had shame back then. You could browbeat them into correcting [mistakes].’
He sees its disruptive potential in the here and now.
Digital Don Juan
“There’s no reason why AI couldn’t be designed to manipulate human beings,” says Reynolds, known for his decades-old Instapundit.com website. “Raw brain power isn’t the best way to do it.”
Yes, the book explores the literal seductive power of an AI-powered device, whether an app, software program, or, eventually, a life-size sexbot coming to a Best Buy near you.
It also shares how manipulative AI can already be and some possible guardrails to prevent it from harming us.
Pop culture already warned us about AI’s seductive power. Think 2013’s “Her,” starring Joaquin Phoenix as a lonely man who falls for a bot voiced by Scarlett Johansson. Or even “The Big Bang Theory,” when the awkward Raj (Kunal Nayyar) falls in love with his Siri device.
“Seductive AI in the crudest sense … is looking more realistic as time passes,” Reynolds says. “You’ve seen these stories. … Women marrying their AI boyfriends. There’s just enough of that out there. You can’t dismiss it as ridiculous.”
The case of the 14-year-old Florida boy who took his own life after sharing suicidal thoughts with an AI bot named after “Game of Thrones” character Daenerys Targaryen is hard to forget.
Blind faith
And it could soon get worse.
“One of my recurring themes in the book … year after year, the machines get better and people stay about the same,” he says, a scary thought given the technological progress we have already seen. “People’s ability to see through this stuff is a flat line.”
Humanity’s wobbly mental health status makes “Seductive AI” fears more profound.
“There’s a large number of people who are losing contact with objective reality. It’s encouraged by social media and a lot of machine affirmation. … The various AI chatbots will basically tell you how smart you are,” he says.
Even some terrible ideas, when fed into an AI bot, will spit back encouraging banter.
“All these platforms … not just the AI ones, foster engagement by pushing various emotions — fear, hatred, sometimes love,” he says.
RELATED: 6 movies that warned us about AI
Sunset Boulevard/Getty Images
The bot stops here
“Seductive AI” offers some possible guardrails, like suggesting that AI firms have a fiduciary duty to the person impacted by their expertise. That could allow people to sue if the bot’s behavior is in breach of that contract.
“The company producing the entity should be held liable for any breaches, exactly as if they had been made by a human employee acting for the company itself,” he writes in the book.
Reynolds says mainstream media outlets have done their part to promote the upside of AI, like fawning press over the rise of self-driving cars.
“Every single story you read in the automotive press was positive,” he says, downplaying the potential for fatal accident. “AI stuff was all super positive for a while. … Now that seems to have faded.”
The Blogfather
Reynolds previously wrote “The Social Media Upheaval” (2019) and “An Army of Davids: How Markets and Technology Empower Ordinary People to Beat Big Media, Big Government, and Other Goliaths” (2007).
He’s best known in conservative circles for Instapundit.com, an old-school site with constantly updated links to the latest news and commentary. He was part of the early blogging wave that challenged mainstream media, with some stunning successes. In fact, he was so influential on other DIY pundits that he earned the nickname the Blogfather.
“The media actually had shame back then,” he says. “You could browbeat them into correcting [mistakes].” Take Dan Rather’s National Guard story, in which the CBS anchor claimed President George W. Bush shirked his duties based on manufactured evidence. The story might have stood unchallenged if not for several citizen journalists like the team behind Powerlineblog.com.
A simpler time
And he has his “beefs” with the current right-leaning media landscape. He recalls a simpler time in the digital arena.
“The period of 2004 to 2008 was kind of a golden age of independent media, before the walled gardens of Facebook and other platforms took over,” he says. It helped that journalists took criticism more seriously at the time.
The early blogging days also saw friendlier ties between left- and right-leaning bloggers. Now, that sense of brotherhood is gone, he says.
“It’s hard to have a civil discussion about anything now,” he says. “It’s a very unhealthy environment.”
As for his latest project, he admits the alluring nature of this technology boils down to something elemental.
“Yes, AI is extremely useful,” he says. “That’s another way of being seductive.”
Read the full article here
