Unpacking My Physical AI Portfolio

Gaetano · Gaetano · 09 августа 2026, 19:29 · ⏱ 7 мин чтения  | Читать в Substack ↗
Резюме
The author argues that the bottleneck for physical AI and robotics is shifting toward high-precision, real-world training data, positioning infrastructure providers as critical beneficiaries. As robotics companies scale up to industrial-level data pipelines requiring millions of hours of first-person video, companies that can build and license these specialized datasets will see significant margin expansion and recurring revenue.
  • •The author has allocated 68% of their intended capital across 14 companies in a dedicated 'Physical AI' portfolio.
  • •INOD reported Q2 revenue of $92.1M, up 58% YoY, with adjusted gross margins improving from 47% to 49%.
  • •INOD's customer concentration improved, with its largest customer dropping from 56% to 37% of revenue, while a second tech customer grew to 34%.
  • •INOD is scoping a massive 2-million-hour first-person data collection program for a leading robotics company, equivalent to roughly 1,000 person-years of full-time data collection.
  • •INOD developed a drone-detection AI model that beat previous benchmarks by 6.45 percentage points, demonstrating capabilities beyond basic data labeling.
  • •Rahul Singhal will become President and CEO of INOD effective September 30, with current CEO Jack Abuhoff moving to Executive Chairman.
Время чтения 7 мин
Объём 7,877 симв.
Категория finance
Идеи
crux_capital_ Independent Photonics & AI Infrastructure Researcher
The company is successfully transitioning from basic data labeling to providing specialized, industrial-scale physical AI training data infrastructure for robotics, evidenced by 58% YoY revenue growth
The company is successfully transitioning from basic data labeling to providing specialized, industrial-scale physical AI training data infrastructure for robotics, evidenced by 58% YoY revenue growth and improving gross margins.
INOD Выс.
crux_capital_ Independent Photonics & AI Infrastructure Researcher
The stock is a core holding in the author's Physical AI portfolio and recently demonstrated strong post-earnings price momentum.
AEVA Выс.
crux_capital_ Independent Photonics & AI Infrastructure Researcher
Held as part of the author's 14-company Physical AI portfolio, showing positive post-earnings momentum.
BKSY INDI Выс.
crux_capital_ Independent Photonics & AI Infrastructure Researcher
The massive scale of physical AI data collection (e.g., 2 million hours of first-person video) will require unprecedented compute for multimodal model training and simulation, directly benefiting Nvid
The massive scale of physical AI data collection (e.g., 2 million hours of first-person video) will require unprecedented compute for multimodal model training and simulation, directly benefiting Nvidia's robotics (Isaac/GR00T) and datacenter segments. Risk: Data collection bottlenecks could delay the timeline for physical AI model training and subsequent compute demand.
NVDA
crux_capital_ Independent Photonics & AI Infrastructure Researcher
The article notes 'leading robotics companies' are piloting industrial-scale first-person data collection pipelines. Tesla's Optimus program is a primary driver of this demand for human motion and man
The article notes 'leading robotics companies' are piloting industrial-scale first-person data collection pipelines. Tesla's Optimus program is a primary driver of this demand for human motion and manipulation data. Risk: High costs of physical data collection compared to synthetic data generation could pressure margins for early robotics developers.
TSLA
Ещё от Gaetano
$AAOI Earnings. They Said Some Wild Stuff.
07 авг, 03:49
$AAOI The Time Has Come
06 авг, 13:27
VIAV, VECO, AEVA, VPG Reported Earnings Today. One New Position. One Big Shock.
06 авг, 05:17
Trump’s Chinese Optical Proposal Is A Really Big Deal. Please Read.
05 авг, 12:38
$VIAV Reports Tomorrow. Will It Follow Suit?
05 авг, 03:23

This newsletter, published August 09, 2026, features crux_capital_ discussing INOD, AEVA, BKSY, INDI, NVDA, TSLA. 5 trade ideas extracted by AI with direction and confidence scoring.

Speakers: crux_capital_  · Tickers: INOD, AEVA, BKSY, INDI, NVDA, TSLA