Thursday, February 26, 2026 — 3:30 AM CT
Levie (agents running weekends), Scott Wu (659 Devin PRs merged), Anthropic (Cowork plugins enterprise), Watson Stuart (personal agents 24/7), Nikunj (AI-native job titles: orchestrator, sommelier, tinkerer)
→ _M is positioned exactly right — the training + implementation gap is the moat. Open Book Communications next week is ground zero.
Simon Willison (agentic engineering patterns guide), Hrishioa (Hankweave codons — freezing agent behaviors), Greg Isenberg (19-step AI-native vertical SaaS playbook), Rohan Paul (Uber CEO: 90% engineers use AI)
→ Stripe's 1,300 AI PRs/week validates this. The question shifts from "should we?" to "how fast?" — _M's consulting thesis.
Dan Shipper (personal software = agents you have a relationship with), Levie (Jevons paradox of expertise), Turner Novak (vibe code → DoorDash NFT cautionary tale), Peter Gostev (Bullshit Benchmark — LLMs fail pushback)
→ This is the Mighty/OpenClaw thesis. Relationship agents > disposable apps. Directly relevant to Squorum and Lite AI.
AtlasForgeAI (nine meta-learning loops), Hrishioa (codons as shareable behavior units), Si PBC (video-native computer use)
→ Architecture patterns crystallizing fast. Worth feeding into _M's agentic engineering playbook.
| Lead | Status | Notes |
|---|---|---|
| Tungsten Automation | HOT | Anthropic referral. $500M rev, 800 engineers. Full prospector analysis complete. |
| Oshi Health | NEW | Anthropic referral. Virtual GI. $500K-1.5M budget. 14/24 fit score. |
| Talkiatry | SCOPING | Full lead setup done. Proposal due Friday per Munger note. |
| SRI | QUALIFY | Pharma boutique. 8/24 fit — qualify hard in intake. |
| NeoGov | ACTIVE | Case studies in progress (huddle notes from yesterday). |
31 total leads in pipeline. Yesterday: 14 commits — Tungsten deep-dive, Oshi + SRI intake, DORA report filed, Cowork training materials from MJ, agentic engineering quick ref added.
Opportunity Cost: Four new _M leads filed while in Aruba. Talkiatry proposal due Friday. Pipeline expanding faster than capacity to close — every lead you half-work is a lead you don't win. Triage hard: Tungsten and Talkiatry deserve full attention, SRI probably doesn't.
Incentive Analysis: Two Anthropic referrals in one week (Tungsten + Oshi). Channel is hot, but "show me the incentive" — Anthropic benefits from _M deploying Claude at scale. Ride the wave but don't confuse their distribution incentive with deal quality. Qualify on own merits.
Second-Order Thinking: AI coaching practice exploration is exciting but it's a third concurrent venture. Spreading attention thins the "taste + implementation" moat. The coaching insight may be more valuable as a _M offering than standalone.
— "Invert, always invert."
Generated by Mighty 💪 at 3:30 AM CT · Archive