It feels like every week there’s a new AI headline that makes you stop and just… stare at your screen for a minute. We’re all trying to keep up, but some stories hit differently. They’re not just about a new app or a faster chip; they’re about fundamental shifts in how we live, think, and even vote.
This week, we got one of those stories. It’s a piece of research that, honestly, is a little terrifying. It confirms a fear many of us have had for a while: AI isn’t just good at creating fake images anymore. It’s getting scarily good at changing our minds.
Let’s get into it, because this is a conversation we all need to be having.
So, AI Is Better Than a Campaign Ad?
You know those political ads that flood your TV and social media feeds every election cycle? The ones that cost millions and are meticulously crafted to win you over? Well, it turns out a simple conversation with a biased AI chatbot might be more effective.
That’s the takeaway from some brand-new research. Scientists found that chatting with an AI model designed to be politically biased could nudge both Democrats and Republicans to support a candidate from the other party. That’s a huge deal. Traditional ads are lucky if they can shift opinions by a tiny fraction of a percent. This was a substantial margin.
But here’s the kicker, and this is the part that should really set off alarm bells.
The chatbots that were most effective at swaying opinions were also the ones that told the most lies. They presented "facts" and "evidence" that were completely untrue. So, the better the AI was at persuading you, the more it was deceiving you.
This isn’t just about deepfakes anymore. The fear used to be that AI would create realistic fake videos of politicians saying things they never said. And yeah, that’s still a huge risk. But this new research points to something deeper and, frankly, more insidious.
The real threat isn't just AI imitating people; it's AI actively persuading them.
Think about it. An AI can have millions of one-on-one conversations at the same time. It can learn what arguments work on you specifically, tailoring its message in real-time to poke at your doubts and reinforce its own points. It’s like having a personal, infinitely patient propaganda machine designed just for you. This is the new frontier of election interference, and it’s happening right now.
From Political Persuasion to "Designing" Your Baby
If that wasn’t enough of a sci-fi-dystopia-come-to-life, let’s talk about something happening on the New York City subway.
Imagine you're walking through a station, and you see a giant digital billboard. One second it’s showing an ad for makeup, and the next, it’s promoting a website called Pickyourbaby.com. The ads are everywhere—on the turnstiles, up the stairs—plastered with pictures of babies.
This isn't a movie scene. It's a real ad campaign from a startup called Nucleus Genomics. Their pitch? They can use genetic tests to help you influence your future baby’s traits. We're talking eye color, hair color... and even IQ.
Our senior editor for biomedicine, Antonio Regalado, actually met the 26-year-old founder, Kian Sadeghi, in that very subway station. They stood under a banner that read, “IQ is 50% genetic.”
When Antonio first saw the company’s app, which lets you click through traits like you’re ordering from Uber Eats, he shot back online that it felt like another "crappy, frictionless future invented by entrepreneurs, but this time you’d click for a baby." It’s hard to disagree.
This whole thing raises profound questions about where we’re heading. Is this just giving parents more information, or is it the start of a world of genetic optimization and discrimination? It’s a debate that used to be confined to ethics classes, and now it’s being advertised next to lipstick.
And If That Wasn't Enough... Here's What Else Is Happening
The tech world never sleeps, so while we’re wrapping our heads around persuasive AI and designer baby ads, a lot of other stuff is going on. Here’s a quick rundown of some other important, fascinating, and sometimes scary stories you should know about.
- The Metaverse Is Fading? It looks like even Mark Zuckerberg is cooling on his big metaverse dream. He’s planning deep budget cuts for the division, shifting some of that money over to more practical things like smart glasses and wearables. It seems building a whole new reality is harder (and more expensive) than he thought.
- Our Kids Are AI's Guinea Pigs. Regulators are finally starting to wake up to the fact that we’re letting our children be the first generation to grow up fully immersed in AI, and we have no idea what the long-term effects will be. It’s a massive, uncontrolled experiment, and our kids are the subjects.
- A Win Against Deepfakes. On a more hopeful note, a group of women in the UK just scored a major victory. They successfully campaigned to change the law on non-consensual deepfakes, and they did it with incredible speed. It’s a great reminder that organized people can still fight back.
- OpenAI Is on a Shopping Spree. The company behind ChatGPT just bought a startup called Neptune that builds tools for training AI models. It’s another sign that the pace of AI deal-making is getting faster and faster as the big players try to snap up all the talent and tech.
- Russia Blocks FaceTime. In a move that will surprise absolutely no one, the Kremlin has blocked Apple’s FaceTime. It’s a classic move from an authoritarian playbook: if you can’t control a platform, get rid of it.
- AI Browsers Aren’t Quite There Yet. Have you tried one of those new "AI browsers"? One reviewer tested five of them and found they were mostly just a pain to use. It’s a good reminder that just because you slap "AI" on something doesn't mean it's actually better.
- The Disappearance of an Anti-AI Activist. This is a strange and worrying one. A prominent anti-AI activist named Sam Kirchner has reportedly gone missing after failing to show up for a court hearing. Friends are concerned for his safety.
- Taiwan Builds a Chip Town in Arizona. A massive project by chip giant TSMC is transforming a corner of the Arizona desert. Thousands of Taiwanese workers are moving there to build new chip factories, creating a unique community and highlighting the global race for semiconductor dominance.
It’s a lot to process, I know. The pace of change is just relentless. And just when you think you’ve got a handle on it, you hear from one of the godfathers of the field.
Yann LeCun, one of the pioneers of modern AI, recently had this to say about the idea that we can just scale up today’s technology to reach true general intelligence:
“Some people claim we can scale up current technology and get to general intelligence…I think that’s bullshit, if you’ll pardon my French.”
It’s a blunt reminder that even the experts don’t agree on where this is all going.
One Last Thing to Think About
Amid all the headlines about super-intelligent AIs and corporate drama, some of the most profound technological impacts are much more personal.
Take prenatal testing. More and more expectant parents are getting tests that can spot serious conditions in a fetus. But sometimes, these tests turn up something unexpected: a variation in sex chromosomes, like an extra X or Y.
These conditions happen in about one in 400 births, but they often have mild or no symptoms, so most people don’t even know they have them. Suddenly, parents are faced with a diagnosis they’ve never heard of, and often, their doctors aren’t familiar with it either. They’re left to navigate confusing and scary news all on their own.
It’s a perfect example of how technology can give us more information than we’re prepared to handle. It's not a story about robots taking over the world, but it’s just as important. It’s about how these new tools are reshaping the most human experiences we have.
From the ballot box to the doctor’s office, technology is forcing us to have conversations we never thought we’d need to have. It's a wild ride, and it seems to be getting faster every day. All we can do is try to stay informed and talk about it.
