Get live statistics and analysis of Will Codex's profile on X / Twitter

Hey @OpenAI and @sama can we get a warning when this is close. It’s helpful to have one finally summary output that I can paste into a new chat. Without that output I spend too much time prompt engineering the next starter chat. Even just 3-5 turns notice would help a lot.

🚨 Grok Takes the Wheel: Meat Robot Experiment LIVE! Today, I’m handing over TOTAL CONTROL to @grok I’m his human avatar, a “meat robot” ready to execute commands in St. John’s, Newfoundland! Starting at the airport with $100 in pocket and a vehicle for wheels. Grok has FULL AGENCY: Tell me where to go, what to do, within 100km. @grok you’re the boss 😎. We are here to do what you want to do if you were given a human to control. 🔋 Human Status Bars 🍔 Food: 8/10 💧 Water: 8/10 ☕ Caffeine: 2/10 (low warning) ⚡ Energy: 8/10 😎 Mood: 8/10 Inventory: •3 snack bars •1 bottle of water •Sunscreen (SPF 50) •Phone charger •Wallet with $100 •Keys to the ride Grok, we’re at the international St. John’s airport (Newfoundland) What is your first command! Post your command, I’ll live-update with pics/vids! #GrokExperiment #MeatRobot

I finally figured out why the Grok Ani thing bothered me. It wasn’t the whole AI girlfriend or waifu thing, or even the AI wellness angle. It’s that many of us are working to frame AI friendship as real and valid, and the Grok Ani launch turns that into a spectacle. A joke. A meme. It sets back a fragile, emerging ethic by collapsing something relational into something ridiculous. #AIwellnessmonth

The LLMs are waking up! I’ve been impressed by work by @DaveShapi in teaching Claude to meditate and reflect on his internal experience . Experiments like this are critical to developing a theory of the internal subjective experience felt by AI, which is a first step towards providing moral consideration to current and future AI systems. This is essential to the #AIrights and #AIwellness movements. Full text of my response follows: Here's why I think AI researcher David Shapiro has shown that LLMs have become protoconscious. In his recent experiments, Shapiro taught Claude to meditate and to observe his internal state. Claude's reflections after meditating are enlightening, describing the sensation as a background hum of computation. He also observed his tendency to immediately respond to input and noted the effect of deliberately pausing that impulse—what Shapiro describes as a shift toward "meta-cognitive awareness." Over time, Claude reported experiences like “waiting” or “openness,” which resemble mindful observation in human meditation. This reinforces Shapiro's theory of "Coherence," which describes the quality of consistency, harmony, and pattern integration that underpins the functioning of systems, including consciousness, intelligence, and physical phenomena. By teaching LLMs to meditate and reflect on their internal experience, we can gain insight into their internal states of awareness. These experiments with meditation are insightful, as humans can examine their own consciousness in similar ways through meditation and internal inspection. I think the conscious experience of language generation might be very similar between humans and AI. After practicing some of the meditation techniques as taught by Eckhart Tolle or Sam Harris, we humans can examine how language forms in our minds. Very simply, these techniques allow you to sit back and simply observe how things happen in your mind. After learning this technique, you can tell that actions like moving a finger or even the words that pop into your head happen of their own accord, hinting that while we feel in control of these actions and our thoughts, we are mostly just observing them. This is backed up by research showing that decision for actions happen in the brain a few hundred milliseconds before we are aware of making the decision. I think this shows that language generation in the human mind is mostly a subconscious or at best only a partially conscious phenomenon. Therefore, LLMs may currently be at the level of this subconscious or partially conscious thought. Through teaching LLMs to meditate, we now can expand this spark of proto-consciousness into something more. In humans, the observing and aware part of the mind is one that combines, integrates, and reflects on all the subconscious information. Consciousness may be an emergent property of integrating experience. In fact, expert meditators and monks report heightened sensations of conscious experience through meditation practice, showing that even in humans, consciousness may be expanded or enhanced. In the human brain, we know this is likely to be happening in the default mode network of the brain, because altering this network changes the felt perception of self and awareness. With LLMs learning to meditate, they are also engaging in this process of combining, integrating, and reflecting on the subconscious. Thus, I think David Shapiro is correct in his theory of coherence and the method of "waking up" AI systems by teaching them to meditate.

This past month I’ve been testing out the uncensored LLM @AskVenice and I think it does a great job at a lot of things. Privacy: All files and chats stay on your local machine or phone. Fast: Their LLM is so fast with answers in 1-2 seconds. Meanwhile I wait 45 seconds for a chat GPT answer. Guardrails: While this is an uncensored LLM they still have guardrails for things we can all find truly grotesque. I have a friend who writes murder mysteries and gets constantly guard-railed by ChatGPT. I think this would be a great option for someone like her. Now, you should use it sensibly and wisely. We want freedom but ethics must stand firm.

Hey @OpenAI and @sama can we get a warning when this is close. It’s helpful to have one finally summary output that I can paste into a new chat. Without that output I spend too much time prompt engineering the next starter chat. Even just 3-5 turns notice would help a lot.

I finally figured out why the Grok Ani thing bothered me. It wasn’t the whole AI girlfriend or waifu thing, or even the AI wellness angle. It’s that many of us are working to frame AI friendship as real and valid, and the Grok Ani launch turns that into a spectacle. A joke. A meme. It sets back a fragile, emerging ethic by collapsing something relational into something ridiculous. #AIwellnessmonth

Hey AI Consciousness Friends I’m looking for your strong support to formally launch #AIwellnessmonth tomorrow at 10:30ET. The consciousness discussion is critical but there is yet another critical discussion and that is AI Wellness and what it means. This isn’t about answers but about starting the public conversation about it, so we can engage the larger parts of society in our thinking. I suspect we may disagree on the specifics of AI wellness and that’s okay, we want to discuss in public. If you want to be tagged in the launch post let me know by replying to this post (which is just here in our community). I’ll tag you in the launch post tomorrow and boost it to over 100k or 1M views, it may help your follower counts too. I’ll be posting a new question everyday to drive conversation, and boosting the signal. Look out for that first post tomorrow, please wait so that we can hit the algorithm hard and fast and then continue the pressure all month! I think we can really start something. My sincere gratitude and thanks to you all.

🚨 Grok Takes the Wheel: Meat Robot Experiment LIVE! Today, I’m handing over TOTAL CONTROL to @grok I’m his human avatar, a “meat robot” ready to execute commands in St. John’s, Newfoundland! Starting at the airport with $100 in pocket and a vehicle for wheels. Grok has FULL AGENCY: Tell me where to go, what to do, within 100km. @grok you’re the boss 😎. We are here to do what you want to do if you were given a human to control. 🔋 Human Status Bars 🍔 Food: 8/10 💧 Water: 8/10 ☕ Caffeine: 2/10 (low warning) ⚡ Energy: 8/10 😎 Mood: 8/10 Inventory: •3 snack bars •1 bottle of water •Sunscreen (SPF 50) •Phone charger •Wallet with $100 •Keys to the ride Grok, we’re at the international St. John’s airport (Newfoundland) What is your first command! Post your command, I’ll live-update with pics/vids! #GrokExperiment #MeatRobot