I was on the communications & policy team at Google DeepMind from 2018 - 2022. When I first joined GDM, external communication about the possibility of human extinction was not permitted, by anyone, at any level of the organization. If asked about existential risk, researchers were PR trained to respond along the lines of: "It's not useful to engage in that kind of alarmism. Some people confuse AI with movies like Terminator -- that's simply not the reality. The AI we develop will be safe by design. After all, we're building it!" And then steer conversation towards beneficial applications in health, climate, etc. Meanwhile, the internal reality was that AI alignment was not solved, reward hacking was the default behavior of RL agents, and there were far too few people working on the problem. After months of advocacy, the policy was updated: positively valenced, comms-friendly content on AI safety was permitted (like deepmindsafetyresearch.mediu…). Note the positive, nice-sounding way that it says "superintelligence could lead to human extinction." The gap between the internal reality and external communications is closing because the risk/reward has changed, and because the evidence is harder to dismiss now. Not because it's a PR stunt or political psy-op. The truth is being said out loud because RSI is now so imminent that no other option makes sense.

People have been asking how I feel about @AnthropicAI naming their latest model Mythos. I feel good about it. @MythosVentures and @AnthropicAI share many aesthetics and values. It doesn't surprise me that we'd both find meaning in this word. Mythos (μῦθος) means: - tale, story, legend, myth - plot (in Aristotle's Poetics: ἡ τῶν πραγμάτων σύστασις, "the standing-together of the things-done") - a set of beliefs or assumptions about something I named our firm Mythos because building a generational company requires telling a story about the future, over and over again, that few believe or will listen to. This is what the early days of an exponential revenue chart feel like. Ultimately that far-fetched story becomes a reality through foresight, ingenuity, and sheer force of will. The staged release of Claude Mythos marks a turning point in the story of AI. They might have instead called it peripeteia (περιπέτεια) -- "turning point" -- an essential plot element for any good drama. But this doesn't quite have the same ring to it. Mythos is a good name. Names are powerful. Finding the true name of a thing is more an act of discovery than an act of choice. For me, I first learned the word long ago -- my older brother @sachinmaini was always adamant that I study the Greek & Roman classics growing up. I was reminded of it in 2022, thinking about the future, enjoying a particularly delicious Hellenic lager while listening to house music on a beach in Mykonos. If you're American you may not have seen this bottle before (you can't easily get it in the U.S.). But now you know. Neither @MythosVentures nor @AnthropicAI were the first to discover this name. My friends, the singularity is beer.