{"id":"Mgix0ppI0jw","title":"illia's 'billionaire chief of staff' AI Agent, FormulaLab's Deep-Dive on NEAR's TEE & Claude Updates","description":"Join the NEAR Legion at https://nearlegion.com/ \nLegion Events Calendar: https://luma.com/near-legion \nFollow us on X:   / nearlegion  \nIronClaw: https://agent.near.ai/\n\nIllia x Business Insider Interview: https://x.com/BusinessInsider/status/...\n\nFormulaLabs on NEAR's Verifiable Private AI Stack: \nhttps://x.com/BASbyLudium/status/2044...\nClaude Code Routines: https://claude.com/blog/introducing-r...\n\nGemini 3.1 Flash TTS: https://blog.google/innovation-and-ai...","channel_name":"NEAR Protocol","published_at":"2026-05-05T00:00:00.000Z","thumbnail_url":"https://img.youtube.com/vi/Mgix0ppI0jw/hqdefault.jpg","transcript_text":"Good morning, Legion. We are back with the daily video, at least until I have no more time to make them. Plenty of ground to cover today. Ilia's latest interview, a solid deep dive on Near's private AI stack, and some product drops on the AI side worth flagging. Let's dive right into it. Ilia sat down with Business Insider a couple of weeks ago to talk about what daily life with agents actually looks like for him. He's running about a dozen at a single point in time, including one configured with the prompt, quote, \"You're a billionaire chief of staff that pulls together his meeting notes, drive docs, and Slack messages into a weekly executive briefing.\" Others handle news aggregation, market reads, coding, and growth experiments. His bigger point in the interview is that society, the internet, institutions are not ready for AGI to actually show up. The agents that will be handling your travel, your money, your health, they can't depend on a single frontier lab as the gatekeeper. As Ilia put it, this is going to have all your information. Literally, your life will be there. So, you don't want any singular company to have controller access to this. That's exactly what Near AI is being built for, an open-source auditable back-end trust layer. So, the data and execution don't sit behind one company's promise. He also flagged the manipulation risk, pointing at the Grock quote white genocide incident from last year as the kind of thing that happens when one company quietly controls the answers. It's worth a read. There's a link in the description to this article. On the same theme, Formula Labs published a piece on BAS this week, walking through how Near's verifiable private AI stack actually works in production. Near AAI cloud running inference inside TEES, Ironclaw for Asian execution, and the Brave integration as the proof point. Leo AI now runs Near AI's TE's base models with a verifiable privacy label users can check themselves. Also worth a read if you want the deeper technical walkthrough on what the Near AI team is cooking up. Over at Enthropic, Cloud Code just got routines in research preview. You configure a prompt, a repo, and connectors once, and the routine runs on a schedule or in response to events without your laptop having to stay open. Teams are already using it for nightly backlog triage and automated PR reviews. It's available now on Pro, Max, Team, and Enterprise plans. Overall, folks, models keep getting more capable and more agentic, which makes the question Elia is raising only more important. Who controls the data? And can you actually verify it? That's exactly the gap Ironclaw and Near AI are filling. Join the Near Legion today and follow along. Catch you in the next one.","transcript_synced_at":"2026-11-03T04:30:33.800Z","segments":[{"start_seconds":0,"text":"Good morning, Legion. We are back with the daily video, at least until I have no more time to make them. Plenty of ground to cover today. Ilia's latest","url":"https://www.youtube.com/watch?v=Mgix0ppI0jw&t=0s"},{"start_seconds":8,"text":"interview, a solid deep dive on Near's private AI stack, and some product drops on the AI side worth flagging. Let's dive right into it. Ilia","url":"https://www.youtube.com/watch?v=Mgix0ppI0jw&t=8s"},{"start_seconds":16,"text":"sat down with Business Insider a couple of weeks ago to talk about what daily life with agents actually looks like for him. He's running about a dozen","url":"https://www.youtube.com/watch?v=Mgix0ppI0jw&t=16s"},{"start_seconds":23,"text":"at a single point in time, including one configured with the prompt, quote, \"You're a billionaire chief of staff that pulls together his meeting","url":"https://www.youtube.com/watch?v=Mgix0ppI0jw&t=23s"},{"start_seconds":30,"text":"notes, drive docs, and Slack messages into a weekly executive briefing.\" Others handle news aggregation, market reads, coding, and growth experiments.","url":"https://www.youtube.com/watch?v=Mgix0ppI0jw&t=30s"},{"start_seconds":39,"text":"His bigger point in the interview is that society, the internet, institutions are not ready for AGI to actually show up. The agents that will be handling your travel, your","url":"https://www.youtube.com/watch?v=Mgix0ppI0jw&t=39s"},{"start_seconds":47,"text":"money, your health, they can't depend on a single frontier lab as the gatekeeper.","url":"https://www.youtube.com/watch?v=Mgix0ppI0jw&t=47s"},{"start_seconds":51,"text":"As Ilia put it, this is going to have all your information. Literally, your life will be there. So, you don't want any singular company to have","url":"https://www.youtube.com/watch?v=Mgix0ppI0jw&t=51s"},{"start_seconds":58,"text":"controller access to this. That's exactly what Near AI is being built for, an open-source auditable back-end trust layer. So, the data and","url":"https://www.youtube.com/watch?v=Mgix0ppI0jw&t=58s"},{"start_seconds":66,"text":"execution don't sit behind one company's promise. He also flagged the manipulation risk, pointing at the Grock quote white genocide incident from last year as the kind of","url":"https://www.youtube.com/watch?v=Mgix0ppI0jw&t=66s"},{"start_seconds":74,"text":"thing that happens when one company quietly controls the answers.","url":"https://www.youtube.com/watch?v=Mgix0ppI0jw&t=74s"},{"start_seconds":78,"text":"It's worth a read. There's a link in the description to this article. On the same theme, Formula Labs published a piece on BAS this week, walking through how","url":"https://www.youtube.com/watch?v=Mgix0ppI0jw&t=78s"},{"start_seconds":85,"text":"Near's verifiable private AI stack actually works in production. Near AAI cloud running inference inside TEES, Ironclaw for Asian execution, and the Brave integration as the proof point.","url":"https://www.youtube.com/watch?v=Mgix0ppI0jw&t=85s"},{"start_seconds":96,"text":"Leo AI now runs Near AI's TE's base models with a verifiable privacy label users can check themselves. Also worth a read if you want the deeper","url":"https://www.youtube.com/watch?v=Mgix0ppI0jw&t=96s"},{"start_seconds":105,"text":"technical walkthrough on what the Near AI team is cooking up. Over at Enthropic, Cloud Code just got routines in research preview. You","url":"https://www.youtube.com/watch?v=Mgix0ppI0jw&t=105s"},{"start_seconds":112,"text":"configure a prompt, a repo, and connectors once, and the routine runs on a schedule or in response to events without your laptop having to stay open. Teams are already using it","url":"https://www.youtube.com/watch?v=Mgix0ppI0jw&t=112s"},{"start_seconds":120,"text":"for nightly backlog triage and automated PR reviews. It's available now on Pro, Max, Team, and Enterprise plans.","url":"https://www.youtube.com/watch?v=Mgix0ppI0jw&t=120s"},{"start_seconds":127,"text":"Overall, folks, models keep getting more capable and more agentic, which makes the question Elia is raising only more important. Who controls the data? And","url":"https://www.youtube.com/watch?v=Mgix0ppI0jw&t=127s"},{"start_seconds":136,"text":"can you actually verify it? That's exactly the gap Ironclaw and Near AI are filling. Join the Near Legion today and follow along. Catch you in the next one.","url":"https://www.youtube.com/watch?v=Mgix0ppI0jw&t=136s"}],"youtube_url":"https://youtube.com/watch?v=Mgix0ppI0jw","url":"https://near-pedia.xyz/videos/Mgix0ppI0jw"}