OpenAI 퇴사 후 텔레파시 AI 모델 구축을 위해 나선 이유

On Thursday, July 23rd, I resigned from OpenAI. On the 24th, I started as a Founding Researcher at Conduit . We're building telepathy: thought-to-text models, trained on non-invasive neural data. I'll talk about: Thought-to-text: what using telepathic tools will be like in 2027, 2030, and 2035 How do we build thought-to-text: why Conduit is collecting immense quantities of non-invasive neural data Why I joined: I joined because of the audacious vision, the brilliant people, and the fun problems Hi: about me 1. Thought-to-text, aka telepathy My prediction: In a couple years, the main way we'll talk with our AIs is with our thoughts. They will not be a super-smart automated intern or coworker that you desperately try to keep up with. They will be a natural, joyful extension of you. I wrote down vignettes of this future below. I view them as optimistic but highly plausible. 2027 2027 — A morning at Conduit I put a band around my head, and open my laptop at Conduit. It’s 9 am. My device pairs with my laptop over Bluetooth. I open, without loss of generality, Codex. I think about the day's work as I look at my code, review yesterday's notes, and prepare for our 9:30 am standup. My GPT-7 agent has been chugging away at the new encoder I’ve been exploring for over a day now. I look at its work. I become confused by the plot labels, annoyed at the AI-speak in its first paragraph, and curious about the symbols in the equations. My vague thoughts get sent to Conduit's model, which uses its priors over language and the kinds of things I might say to output: conduit://thought-stream → codex $ decode --source=neural-latent --autosend --to=codex thought_01 > I’m looking at your plots. Please relabel the legend, and add a text cell below with a simple description of each line. thought_02 > I understand your 2nd paragraph of discussion, but please rewrite the first paragraph using the humanizer skill. thought_03 > Make a textbox explaining the equations with cleaner notation. streaming > █ The plot, paragraphs, and notation each took me 10 seconds to glance over. I have auto-send on, so after each 10-second chunk, the Conduit model deciphers my thoughts and sends it to GPT-7, which assigns each incoming task to a subagent. I look at the new plot, paragraph, and equations. I wonder whether spherical harmonics might be useful after all; Codex spins off a subagent. Convinced by the plot, I decide to make a quick slideshow with a prettified version showing only the baseline and top two lines, to show my teammates; Codex spins off a subagent. I make myself coffee. My thoughts are mostly empty, but I briefly recall what I want to say to the candidate I’m getting lunch with tomorrow. In the background, Codex is prompted to think about info I might want for the lunch chat, and decides that updating the synthetic data scaling plot before showing it off would be prudent; Codex spins off a subagent. I'm not saying words really loudly in my head while getting coffee. I just read the plots as I normally do, and make coffee as I normally do. It feels like magic. 2030 2030 — Industry adapts and Conduit expands The AI companies now train their models to directly interface with Conduit’s latent representations. My encoded thoughts get sent directly to Codex, rather than having to pass through Conduit's decoder model first. That means I easily communicate thoughts that are hard to describe in text, like mental images. I rarely go band-less when chatting with AIs these days. It's annoying, honestly, to go without. With the neural headband, I feel like I have superhuman powers over my laptop! Without it, I feel like I'm talking with a superpowered alien who's trying its best to be helpful but isn't sure what I want and is scared I'll get mad if it does something I don't want. Ugh. Conduit continues to iterate on non-invasive read, but we’ve spun up two new efforts: invasive general read, and general write. Most people are happy to stick with their neural bands, but a good number are excited to get higher fidelity reads via invasive tech. More interesting is the recent excitement in writes. I feel superhuman in my control over my laptop. But my senses are still merely human. It's like if I could control my arms, but I'd lost all feeling in them. Yeah, I can still see my arms, and it's way better than not having arms, but it's still really odd. Like Ian Waterman . I want to feel what my Codex feels. Now that Conduit does general read, learning to do general write is many OOMs more data efficient. More importantly, I want to unlock the other applications of write technology. I want to make my brain as efficient and neuroplastic as when I was 18. 2035 2035 — The AI is no longer “ other ” My AI is a natural extension of me. It feels like a sixth sense and another limb. I wonder about a problem, feel as though I’m literally surfing the web, see glimpses of the websites, get flashes of intuition about the problem, and ultimately derive the answer. It feels fun! My brain is like the Flash. Because of write tech, my thinking is the fastest it's ever been even when I turn my AI off. How superhuman 2035 looks depends directly on how superhuman we, as society, decide to make our AIs. Perhaps we choose to pace ourselves . But I like that this is literally a human-in-the-loop vision of the future, where AI directly empowers humans rather than replacing us. A few quotes about the future of thought-to-AI Writers and speakers more elegant than I have elaborated on this thought-to-AI interface over the past decade. I will link to them, along with brief excerpts. “If we achieve tight symbiosis, the AI wouldn’t be “other”—it would be you and with a relationship to your cortex analogous to the relationship your cortex has with your limbic system.” – Elon Musk, 2017, in the Wait But Why interview that inspired one of Conduit’s cofounders to go into neurotech! “We could plug electrodes into our brains…. I think a merge is probably our best-case scenario. If two different species both want the same thing and only one can have it—in this case, to be the dominant species on the planet and beyond—they are going to have conflict.” – Sam Altman, 2017, in Sam’s personal blog . "Can we translate thought into basic actions?... I want to see what my schedule is today. I want to see what the weather is today. I want to turn the bedroom lights... If you can think, 'I want to turn the lights downstairs off,' and they will turn off, that might feel shockingly like telepathy." – Dean Ball, 2024, in an interview . “It is not possible to understand the long-term future of artificial intelligence without understanding brain-computer interfaces.” – Rob Toews, 2025, in The Next Frontier For AI Is The Human Brain Forbes Article . Quotes Conduit’s co-founders, Rio and Clem. <3 2. How do we build thought-to-text? In theory, it’s simple. Our input is brain activity, and our target output is what the person was doing at the time – for example, what text the person wrote. Given the brain activity, we want to predict output that is semantically similar to what the person wrote. To train models that can predict text given brain signals, we must apply the same lesson learned by those predicting text given speech audio, or text given preceding text: the bitter lesson . The lesson roughly states that you should throw more useful compute at your model, and your model will become better than any ingenious algorithm you could've hand-crafted. That means we must scale up our data collection by orders of magnitude beyond what has ever been done in academia. It’s hard to collect enough data using invasive methods. Few people want a chip in their brain! But non-invasive methods are getting much better . The hardware is improving and getting cheaper, though I apologize for being vague about the particularities of our hardware. As we're training on more data, the model is predicting text that is more semantically similar to the subject-written t