Toshi.Guy

⚫ UNCANNY VALLEY: 94% OF AI AGENTS DO NOTHING - BUT YOUR JOB IS STILL ON THE LINE AI agents can’t book meetings, can’t click ads, can’t even admit when they’re wrong. Only 6% of enterprise seats are used. Most agents sit idle—confident, useless, expensive. And yet: entire teams are being restructured around them. Budgets are shifting. Job postings vanish. They’re not replacing workers because AI works—they’re doing it because pretending it does feels safer than admitting AI still weird and broken. CEO of @Youdotcom and Partner at AIX Ventures @RichardSocher joins @OperationDanish to talk about what’s broken now - and the future we’re sleepwalking into. Fridays at 4:20PM ET. Only on 𝕏 00:41 – Why AI won’t kill software—but it will change search forever. 02:41 – “Google is getting unbundled.” Why users are searching somewhere else. 03:53 – Richard’s real focus: powering complex, work-related queries. 05:33 – Why most companies buy AI licenses—but no one uses them. 06:50 – The next skill: managing AI like a team member. 08:45 – Knowledge agents work. Action agents? Still too risky. 09:53 – Why AI can’t book your family trip—yet. 11:09 – Overconfidence is killing trust. Agents need to learn to say “I don’t know.” 13:20 – Intelligence is costly. Most users don’t want to pay for it. 14:33 – Sometimes a button beats a chatbot. Here’s why. 15:29 – Richard’s go-to agent: a research tool that scans 400+ sources. 18:41 – Agents won’t click ads. That could break the web’s business model. 19:41 – What makes Richard’s agent better? Accuracy and real citations. 21:45 – “Search quality makes or breaks agents.” 23:53 – Reverse RAG: how they trace every claim back to its source. 24:50 – Most AI fails on hype. This one’s winning with customers. 26:38 – Every use case has its own hype cycle—agents included. 28:44 – Coding agents are impressive—but still break under pressure. 31:06 – The real moat? Distribution, data, and depth. 33:37 – Regulated industries need a human in the loop—for now. 34:26 – AI nurses are real—but it starts with logistics, not diagnosis. 35:23 – Final take: what’s next in the agent arms race.

⚫ UNCANNY VALLEY: FLYING CARS ARE REAL—BUT YOU STILL THINK THEY’RE SCI-FI Everyone loves the idea of flying cars—until they see one. And then? The disbelief kicks in. The safety questions. The “not in my city” reflex. But they’re not just real. They’re flying. Legal in multiple countries. Quieter than traffic. Safer than your sedan. In this episode, @OperationDanish sits down with Jim Dukhovny, founder of Alef and creator of the world’s first street-legal flying car. They break down the tech, the trust gap, and why America may fall behind if it doesn’t get out of its own way. Fridays at 4:20PM ET. Only on 𝕏 00:23 – Flying cars, eVTOLs, drones: What’s what—and why it matters 01:59 – China’s legal lead: Behind on tech, ahead on policy 04:19 – “Flying cars” are mostly hype—except this one 06:18 – Why vertical takeoff is a non-negotiable feature 08:08 – Not a drone. Not a toy. A real, road-legal machine 10:14 – The path to affordability: From Roadster to Corolla 11:57 – How it’s simpler—and stronger—than today’s cars 14:16 – Inside the propulsion system that makes it fly 17:32 – Why hardware is back in Silicon Valley 23:45 – Software still powers flight—and scale 25:12 – What happens if @ElonMusk enters the space? 26:42 – Why real competition helps legitimize the category 29:26 – FAA vs. China: Who clears the skies first? 31:16 – Where it’s already legal—and how rollout begins 35:34 – Can you ride one today? Technically, yes 37:18 – It’s 4–16x safer than your car. Here’s the math 38:26 – Why the biggest barrier isn’t tech—it’s trust