Born

Thesis

Born

  • Shift: Perspective stood up. Two same-day user-flagged robotics baskets — mkfilko (substantive stack-map: OS/sensors/brain/body, real financials, honest caveats) and mmatters22596 (engagement-radar "Jensen told you to buy robotics, 10x potential") — plus a 6-month-old @RealJGBanks growth list that incidentally tagged RR/SERV. The convergence of two independent robotics pushes in one day is the
  • Shift: Perspective stood up. Two same-day user-flagged robotics baskets — mkfilko (substantive stack-map: OS/sensors/brain/body, real financials, honest caveats) and mmatters22596 (engagement-radar "Jensen told you to buy robotics, 10x potential") — plus a 6-month-old @RealJGBanks growth list that incidentally tagged RR/SERV. The convergence of two independent robotics pushes in one day is the birth signal: physical AI / robotics went consensus-viral this week. User asked to "build up a robotics perspective."
  • Frame chosen (physics-down, not cheerleading): the basket bifurcates into profitable arms-dealers (compute brain NVDA, OS BB/QNX, vision CGNX/AMBA, cobots TER/ROK/ABB/FANUY, surgical ISRG, warehouse SYM) vs pre-revenue moonshots (humanoid deSPACs CCXI/RR/SERV/XPEV, lidar OUST/AEVA/LAZR/HSAI/INVZ). Thesis = own the layer that sells to every robot and is already profitable; the moonshots are options on an unknown winner. Same picks-and-shovels spine as ai-power-bottleneck.
  • Effect on thesis: status: active, priority: medium. Honest posture — the real book holds none of the spec; only the profitable layer-leaders are trend-hold-eligible, and the most interesting setup is quality-on-sale (NVDA RSI 40 / ISRG RSI 38 pullbacks), NOT the parabolic deSPAC / penny-lidar corner the hype points at. NVDA is paper/research only (real-trade blocklist).
  • Seeded artifacts: robotics watchlist (21 names) created same day; source profiles mkfilko (with the basket captured) + mmatters22596.md; both trial-followed in FOLLOWS.md.
  • Data (06-25 close) confirms the bifurcation: spec/momentum bid (BB RSI 69 / CCXI RSI 96), arms-dealers near highs but calm (TER 64 / ROK 61 / CGNX 57), blue-chips pulled back not broken (NVDA 40 / ISRG 38 / TSLA 39), moonshot tail wreckage (INVZ $0.57 / RR $1.93 / SERV −68%).
  • Folded in same-day (2026-06-27): three pre-existing inputs/ captures wired into the perspective (affects.perspectives) — SemiAnalysis "levels of autonomy" (2025-07-30, structural driver), @NuttyCLD humanoid-through-the-component-layer (2026-05-21, arms-dealer thesis a month early), @ZaStocks Tesla+SpaceX mega-vertical (2026-05-22, cross-links spacex-ipo-liquidity-event). Shows the structural setup + arms-dealer lens predate this week's hype by ~10mo / ~1mo. Also added the wreckage-vs-correction-vs-bad-companies read (user Q): only NVDA/ISRG/TSLA are corrections; lidar tail = structurally-challenged serial-diluters (tech real, business model isn't; LAZR didn't resolve on fetch — verify corporate action); micro-cap humanoids = unproven spec. None is a trend-hold dip.
  • Discord fold-in (2026-06-27, DONE): pulled 8 months of Dumb Money #🦾humanoid (359 msgs, Oct 2025→Jun 2026) → 2026-06-27-dumb-money-humanoid-channel-discord-pull. Added the component/actuator chokepoint layer (the watchlist's real gap): Harmonic Drive Systems (6324.T) + Nabtesco (6268.T) reducer duo "TSMC/ASML of humanoids," VPG hands-sensors (3 NDA'd humanoid customers), Daido Steel (5471.T) magnets, RRX motion. External validation of the arms-dealer frame (JC Parets "robot ETFs are the wrong vehicle"; "NVDA/CGNX = the shovels"). Ref to pull: Morgan Stanley "Humanoid 100" value-chain map.
  • Open follow-ups:
    • Coverage build-out DONE (user-authorized 2026-06-27): added 3 _securities.json family rows — harmonic-drive-systems (6324.T, research-only, no US proxy), nabtesco (NCTKF OTC / 6268.T), daido-steel (DASTF OTC / 5471.T). Added 6 names to robotics (now 27): MVIS, RRX, MBLY, NCTKF, DASTF, 6324.T. Data note: the chokepoint trio (Harmonic Drive RSI 49 / Nabtesco 50 / RRX 59) is un-extended — NOT bid yet (vs the spec tail) — exactly where the durable value should sit.
    • Pulled the Morgan Stanley "Humanoid 100" (2026-06-27-ms-humanoid-100-value-chain) — confirms Harmonic Drive + Nabtesco as the reducer chokepoint and surfaces US-tradeable arms-dealers we missed. Next coverage candidates (user-gated, the durable end): RBC, TKR (bearings), MP (rare-earth magnets), TDY, ADI, ST (sensors), APH, TEL (wiring).
    • ABB + LAZR data-pending (clean 404 on first fetch; retry next fetch:all).
    • Coverage-depth candidates if the theme proves out: China humanoid (UBTECH/Unitree — private/HK), med-robotics (PRCT/STXS), motion/actuators (NJDCY/NOVT).
    • First real test: does the quality-on-sale corner (NVDA/ISRG) turn on volume, or does the spec tail keep leading (= late-distribution, fade)?
4 events
DateTypeEvent
Jun 30 Market SignalRobotics Rotation Source Convergence Jul 2 Tweethis 5 conviction Physical AI names (40-50% of book): OUST/AMBA/CGNX/AEVA/VPG/CCXI Jul 3 Peer ReviewSystematic peer review: crux_capital physical-AI conviction basket vs our engine Jul 8 ArticleThe Unwind Is Washing Out Our Picks-and-Shovels Lanes While Their Fundamentals Inflect Up

No direct external sources are attached to this read.