Latest / 7 Minute Books: Free Book Summaries / Evil Robots, Killer Computers, and Other Myths by Steven Shwartz - Book Summary

Transcript

Evil Robots, Killer Computers, and Other Myths by Steven Shwartz - Book Summary · 7 Minute Books: Free Book Summaries

• 0:01
  Sam: Hey there, welcome back to 7 Minute Books. I'm Sam, and today we're talking about Steven Shwartz's book, Evil Robots, Killer Computers, and and Other Myths. Sophie, I have to ask, did this book actually make you less scared of AI, or was it just me?
• 0:17
  Sophie: Hi there Sam, it's great to be here. And honestly, yes, it really did. I mean, I went in thinking I'd get a lot of technical jargon, but instead it felt like a really sane conversation about what AI actually is and isn't. Shwartz has this way of cutting through the hype without being dismissive.
• 0:37
  Sam: Right, exactly. He's a computer scientist who's actually built these systems, so he's not just guessing. And his main point is pretty simple, we've been telling ourselves the wrong story about AI for decades.
• 0:50
  Sophie: Yeah, the book starts by tracing that story all the way back to ancient Greece. Mechanical servants rebelling, Mary Shelley's Frankenstein, the word 'robot' itself coming from a Czech word for forced labor. These myths tap into deep anxieties, but they're not predictions.
• 1:08
  Sam: And that's the key distinction, right? What we actually have is narrow AI. It's amazing at one thing, like recommending shows or filtering spam, but it can't generalize at all. A chess AI can't make you breakfast. A translation system can't drive a car.
• 1:26
  Sophie: Right. And then there's general AI, or AGI, which would be a machine that can think and learn across any domain, like a human. Shwartz argues we're nowhere close to that. We don't even understand how our own general intelligence works, so how could we program it?
• 1:43
  Sam: That part really stuck with me. He takes on this idea that AI is on an exponential trajectory toward superintelligence. You know, the Kurzweil and Bostrom narrative. And he just systematically dismantles it.
• 1:58
  Sophie: He does. He points out that it assumes intelligence is a single scale you can just crank up, but it might be more like a landscape with trade-offs. Plus, we've made almost no progress toward AGI in decades. And it assumes a superintelligent AI would naturally want to harm us, which is just projecting human motives onto a machine.
• 2:20
  Sam: Yeah, that's a good point. I also really appreciated what he said about job loss. He's not some Pollyanna saying no one will lose their job. He acknowledges automation will eliminate some roles, just like it always has. But he points out that new jobs always emerge that we couldn't have imagined.
• 2:38
  Sophie: The Luddites example is great. They smashed textile machines, and sure, some jobs disappeared, but then entirely new industries were born. Shwartz says the jobs most at risk are routine, predictable ones. Jobs requiring empathy, creativity, complex problem-solving are much safer.
• 2:59
  Sam: Okay, but here's where the book really got me, he talks about the real risks of AI, and they're not robot uprisings. They're things like bias in machine learning. If you train a hiring algorithm on data from a company that historically favored men, it'll learn to discriminate.
• 3:17
  Sophie: Exactly. And predictive policing systems can perpetuate racial bias. These are real, present problems. Plus authoritarian governments using facial recognition for surveillance. These are the things we should be worried about and regulating.
• 3:34
  Sam: Right. And he also addresses the whole 'paperclip maximizer' thought experiment, where an AI turns everything into paperclips. He says that's based on a misunderstanding. Current AI doesn't have goals like humans do. It has narrow objective functions. A chess AI doesn't want to win; it's just designed to select the best move within the rules.
• 3:57
  Sophie: Yeah, the idea that such a system could somehow generalize its objective to the entire universe is pure science fiction. And Shwartz is really good at drawing that line between science and science fiction throughout the book.
• 4:11
  Sam: He's also humble about what we don't know. He admits we don't understand how the brain works, and we don't know how to replicate it. The progress we've made is from brute force computation and massive data, not from deep understanding of intelligence.
• 4:27
  Sophie: That humility is refreshing. He's critical of the AI hype cycle too, where researchers and companies make grandiose claims to get funding. That hype fuels public fear when the promises fall short.
• 4:40
  Sam: So what's the takeaway? For me, it's that we need to stop being afraid of imaginary monsters and start dealing with real challenges like bias, privacy, and the concentration of power. The robots aren't coming to destroy us; they're coming to help, if we let them.
• 4:57
  Sophie: And if you want to go deeper into Shwartz's arguments or explore other books like this, the whole library is over at 7minutebooks.com/app. There are over 6,000 fiction and nonfiction titles you can read or listen to in any language, and it starts at $2.99 a month, $9.99 a year, or $19.99 for lifetime access.
• 5:24
  Sam: So we should worry less about Skynet and more about making sure AI is fair and transparent. That's a message I can get behind.
• 5:32
  Sophie: Absolutely. The book's whole point is that the future of AI will be shaped by the choices we make today, and those choices are better when they're based on facts, not fantasies. We'll see you in the next one.