I do not understand this debate. But since everyone involved seems to love thought experiments, let’s try this one on. Imagine you could connect a fresh “human” form factor brain to a computer. It has never been attached to a human body. - No prior experience. - No childhood. - No memories. - It has no sense of smell, taste or touch. - Its sense of sight & hearing are limited & high friction - It cannot feel physical pain. - Without the ability to experience physical sensation, its ability to experience any type of emotional feelings are at the very least dramatically altered, if possible at all - It cannot die. - It is simply receiving input & responding or it is inactive It is then given the ability to read text on the internet, and by happenstance it finally manages to organize itself to understand what it has read and respond to questions. Does that biological neural net, that happens to share a form factor with a human’s brain, have a consciousness that is in any way relatable to you and I? Obviously not. That’s why the idea that modern LLM’s should be treated with the same standing that we otherwise reserve for other humans is absolutely ridiculous. It’s the same basic fallacy as the claim that discarding an embryo during IVF is tantamount to murder. No. It obviously is not the same thing.

CEO of @anduril, @PalmerLuckey, shows off new autonomous drone systems that will end wildfires to @PeterDiamandis. First, reject the false alarms: "who's the guy smoking a cigarette?... what's a campfire? What's a road flare?" "It's about detection, classification, rejection, match making, and then going and hitting the target." Link to full video👇

CEO of @anduril, @PalmerLuckey, told @PeterDiamandis why his new $10M XPRIZE isn't about AI. "I think that the artificial intelligence side of models is pretty well covered by plenty of people." "I think we've not been doing enough understanding biological intelligence, particularly beyond people." Link to full video👇

.@PeterDiamandis, of @xprize & @moonshots_pod, asked @PalmerLuckey if building has changed in the age of AI. His answer for founders: you're "still on the hook legally" and "still on the hook for all the interpersonal stuff." "We've got a good 10 years of speed running all of the other parts of the business before they cover that." Hobbyist coders thinking of starting a company: "You are in for a rude awakening." Link to full video 👇

The AI Boom is Exploding with @farzyness Meta’s @Muse and Grok @Bot are viral, delightful, useful AI products. What does this mean for the tech sector and broader society? One thing is for certain. The pace of change is accelerating! 00:00 AI Feels Unreal 01:28 AI Boom And New Agents 02:45 Bubble Or Doom Narrative 03:30 Self Driving And Agent Workflows 05:22 Meta Muse Goes Mainstream 06:43 Agency Becomes The Divider 09:28 Entrepreneurs Vs Everyone Else 12:04 Inequality And Social Backlash 17:53 Transition Plans And Safety Nets 21:27 Civilization Self Corrects 26:15 Optimism Window Through 2028 30:03 Social Media Hangover Effect 34:08 Pockets Of Boom And Bust 35:28 States As Policy Laboratories 38:42 Data Centers And Local Dividends 40:12 AI Tools Go 10X 41:30 Agency Meets LLMs 42:41 MetaMuse Signals Shift 44:16 Rocky Transition Ahead 45:03 Global AI Entrepreneurship 47:11 From Digital To Physical 49:34 Drones Versus Humanoids 50:50 Biohybrid Robotics Case 51:54 Hands Are Hard 56:40 Optimism Versus Doom 01:00:05 Skepticism Drives Better Products 01:02:49 Why Meta Wins Distribution 01:03:52 AWS Overflow Compute Angle 01:05:45 Next Snapchat Of AI 01:06:31 GrokBot Creator Economy 01:08:23 YouTube For Productivity 01:10:33 Side Hustle Bot Boom 01:11:42 Welcome To The Future

Same pattern at L3Harris. “AI is a commodity. It’s all about the data.” They had a country-of-origin workflow on frontier models. Fine-tuned open-source models on their own data. Beat the frontier models in under 48 hours. $23B in sales. Decades of proprietary knowledge. L3 owns the Ontology, the compute, the model. L3 owns the advantage.

Well look what we have here! Grok @bot's directly in the Grok web UI. The walls in the Cursor/Grok Chat/Grok Bot ecosystem are dissolving. It won't be long before we have one seamless platform for @elonmusk's digital AI products. Next up, unified subscriptions. And I can't wait for the screen share option from Grok Voice to work natively inside Grok @bot too.

Why isn't there standard for personal memory servers like there is for MCP? Something that can be hosted in the cloud &/or on a personal device. Something that any agent could interact with it. Something that allows you to keep & compound your own data across agents & services. Something that allows different agents to have different access permissions. The closest thing I'm aware of to this is @garrytan's gbrain, and it does a lot of this, but we need an official standard reference architecture that's easier for normies to get started with. This is definitely something that @elonmusk should have @SpaceXAI build for @grok @bot. Or @finkd should have @AIatMeta build for @Muse. And if the labs don't want to do this, then maybe @dhh can cook something up in @OmarchyLinux.

Tesla just registered another 79 Model Ys for its Robotaxi fleet in Texas. Tesla did announce yesterday that their Unsupervised fleet is a lot bigger now, and we have seen wait times decrease significantly lately.