You’ve probably felt it. That creeping unease when you talk to Siri or ChatGPT and realize: this thing is getting inside my head. Not literally. Not yet. But someone just left OpenAI to build the machine that will.
Naomi Bashkansky announced she’s leaving the world’s most powerful AI lab to work on telepathy. Most people rolled their eyes. Sci-fi hype. Neural lace fantasies. But here’s the truth nobody wants to admit: she’s not building a mind-reading gadget. She’s building the single most dangerous platform control point in human history.
Think about it. AI is accelerating faster than any interface can keep up. We still type with our fingers. We still speak into microphones. We still wait for machines to process our clunky, slow, error-prone inputs. The bottleneck isn’t intelligence anymore. It’s the human I/O layer.
Bashkansky is betting that the next billion-dollar breakthrough isn’t a better model. It’s a direct neural channel. And she’s right.
But here’s the twist that makes this story terrifying: Telepathy isn’t about connection. It’s about capture.
Every thought you have becomes data. Every fleeting emotion becomes a signal. Every private decision becomes a training example. The company that builds the thought interface doesn’t just improve AI—they become the gatekeeper of human cognition itself.
One commenter on her post asked: Are you specifically focused on non-invasive methods? Seems like you reach a noise barrier which limits things like this.
That noise barrier is the only thing protecting you from a world where your brain is a product. Non-invasive methods (EEG, fMRI, optical) are noisy, imprecise, and slow. Invasive methods (implants, neural dust) are clearer, faster, and far more dangerous. The trade-off is stark: the clearer the signal, the worse the privacy violation.
Bashkansky hasn’t publicly committed to non-invasive. But even if she does, the pressure to go invasive will be immense. Because the market demands fidelity. Investors want signal clarity. And once you crack the non-invasive barrier, the next logical step is direct—because why settle for a blurry thought when you can have a high-resolution one?
This is the moment. The platform shift nobody is talking about. We’re so busy arguing about AI alignment and AGI timelines that we’re missing the real war: whoever owns the input layer owns the future of intelligence.
OpenAI, Google, Meta—they all want your data. But they’re still fighting over what you type and say. Bashkansky is going straight for the source: your brain.
I’m not saying she’s evil. I’m saying the technology is inherently dangerous. Telepathy promises ultimate connection, but it demands surrender of the most private data we have—our thoughts. And once you give that up, there’s no getting it back.
So let’s stop pretending this is a feel-good story about human-AI collaboration. It’s a hostile takeover. And the most unsettling part? You might volunteer for it.
Because the promise is intoxicating: think something, and it happens. No typing. No speaking. No friction. Pure, instant, mind-to-machine flow. Who wouldn’t want that?
But every time you use a thought-interface, you’re training a model on the most intimate parts of yourself. Your fears. Your desires. Your unspoken judgments. You’re building a permanent record of your inner life. And you have no idea who will own that record—or what they’ll do with it.
This isn’t a prediction. It’s already happening. Neuralink has implanted devices. Synchron has a stentrode. Meta is working on wrist-based neural interfaces. The race is on. Bashkansky is just the latest—and smartest—move.
So here’s what I want you to take away: Stop asking whether telepathy is possible. Start asking who gets to read your mind. Because the answer will determine how much freedom you have in the age of AI.
FAQ
Q: Is telepathy really possible with current technology?
A: Yes, but with severe limitations. Non-invasive methods (EEG, fMRI) are noisy and slow. Invasive methods (implants) offer clearer signals but require surgery. The research is advancing rapidly, driven by the 10x improvement in AI's ability to decode neural patterns from noisy data.
Q: What's the practical implication if this technology succeeds?
A: The practical implication is a complete redefinition of human-machine interaction. You won't type, speak, or click—you'll just think. But the dark side is that every thought becomes trainable data. The company that owns the interface becomes the gatekeeper of human cognition, able to influence, predict, and monetize your inner life.
Q: Isn't this just another overhyped tech trend? What's the contrarian take?
A: The contrarian take is that non-invasive telepathy will never achieve the fidelity needed for mainstream adoption, and invasive methods will face insurmountable regulatory and ethical barriers. But even if the technology stalls, the attempt itself shifts the entire industry toward neural interfaces, forcing every major AI company to invest in thought-data pipelines. The real value isn't the product—it's the data monopoly.