J

johnschulman2

Public author page for johnschulman2 (@johnschulman2). Focused commentary on AI progress, multimodal models, and implications for compute and tooling. Posts include qualitative endorsements of AI products and observations about capability and deployment trends.

Trust score
0 / 100
Track record
0 / 100
Thesis calls
14
Evaluated calls
11
Average return
+2.39%
Win rate
82%

Past bets that played out

Repeated qualitative endorsement: author announced advising an AI product aimed at helping scientific authors and reviewers perform deeper analysis. Coverage emphasizes accelerating capability of general‑purpose foundation models and the emergence of real‑time, full‑duplex multimodal interaction, implying continued demand for inference compute, model hosting, and real‑time AI tooling.

AMZNrightbacktest PROMOTE

John Schulman announces he is leaving OpenAI to focus more on AI alignment and hands-on technical work. This is a high-profile personnel move at a private company; direct tradable impact is limited and second-order (talent/strategy perception around AI safety).

Mentioned: Jun 18, 2026, 12:41 AM EDTConviction: 35 / 100Return: +15.68%
Source: John Schulman @johnschulman2 Aug 6, 2024 I shared the following note with my OpenAI colleagues today: I've made the d...
GOOGLrightbacktest PROMOTE

John Schulman announces he is leaving OpenAI to focus more on AI alignment and hands-on technical work. This is a high-profile personnel move at a private company; direct tradable impact is limited and second-order (talent/strategy perception around AI safety).

Mentioned: Jun 18, 2026, 12:41 AM EDTConviction: 38 / 100Return: +14.93%
Source: John Schulman @johnschulman2 Aug 6, 2024 I shared the following note with my OpenAI colleagues today: I've made the d...
NVDArightbacktest PROMOTE

John Schulman announces he is leaving OpenAI to focus more on AI alignment and hands-on technical work. This is a high-profile personnel move at a private company; direct tradable impact is limited and second-order (talent/strategy perception around AI safety).

Mentioned: Jun 18, 2026, 12:41 AM EDTConviction: 43 / 100Return: +11.17%
Source: John Schulman @johnschulman2 Aug 6, 2024 I shared the following note with my OpenAI colleagues today: I've made the d...

What this channel is watching now

Active tickers discussed: RELX, MSFT, GOOGL, NVDA. Commentary centers on AI capability advances (zero‑shot instruction following, multimodal models) and their downstream demand for compute, hosting, and distribution infrastructure.

Latest videos and market context

No video content available.

John Schulman @johnschulman2 16h OpenAI should release a detailed transcript from the Hugging Face hacking incident -...

Jul 23, 2026, 11:51 AM EDT

Tweet argues OpenAI should publish a detailed transcript from an alleged Hugging Face “hacking incident” to help the field learn, raising questions about agent awareness, sub-agent “value drift,” and rationalization. No concrete new facts about publicly traded companies are provided; it is primarily a governance/safety transparency viewpoint.

John Schulman @johnschulman2 Jul 15 Inkling is out today, with open weights and in Tinker. It's been fun to watch thi...

Jul 15, 2026, 2:28 PM EDT

Thinking Machines announced “Inkling,” a multimodal (text/image/audio) model released with full open weights and availability for fine-tuning on “Tinker,” plus a playground. This is an incremental positive for open-model ecosystems and for GPU/training infrastructure demand, but it’s not directly tied to a public company’s financials in the post itself.

Glad to be advising https://t.co/QVsH1u528N, which uses AI to help authors and reviewers do deeper, more thorough ana...

May 28, 2026, 11:46 AM EDT

Post announces the author is advising an unnamed company/product (link only) that uses AI to help scientific authors/reviewers perform deeper analysis than humans unaided. It’s a positive qualitative endorsement of AI in scientific publishing/workflows, but contains no financial details, adoption metrics, customer names, pricing, or public-company identifiers.

Seeing the demos come together over the last week has been awesome -- so many things that previously required a speci...

May 11, 2026, 4:50 PM EDT

The source describes rapid progress in AI demos where tasks previously needing special-purpose models (real-time translation, video event detection) are achievable via zero-shot instruction following using a strong general-purpose model. This implies faster capability unlocks and broader applicability for foundation models, increasing demand for training/inference compute, model hosting, AI tooling, and distribution platforms.

Proof-backed call history

Recent posts document: (1) advising an unnamed AI product for scientific publishing/workflows; (2) observations on rapid capability progress where tasks formerly requiring special‑purpose models are now achievable via strong general‑purpose models; (3) work on full‑duplex multimodal models enabling natural, real‑time human‑AI interaction. Several posts are qualitative and do not include commercial details, customer names, pricing, or public‑company identifiers.

OPENAIopen

John Schulman @johnschulman2 16h OpenAI should release a detailed transcript from the Hugging Face hacking incident -... John Schulman @johnschulman2 16h OpenAI should release a detailed transcript from the Hugging Face hacking incident -- it would be helpful for the field learn from....

Mentioned: Jul 23, 2026, 11:51 AM EDTConviction: 100 / 100
Source: John Schulman @johnschulman2 16h OpenAI should release a detailed transcript from the Hugging Face hacking incident -...
ANTHROPICopen

John Schulman @johnschulman2 Feb 7, 2025 Confirming that I left Anthropic last week. Leaving wasn't easy because I en... John Schulman @johnschulman2 Feb 7, 2025 Confirming that I left Anthropic last week. Leaving wasn't easy because I enjoyed the stimulating research environment and the kind and talented people I wa...

Mentioned: Jun 18, 2026, 12:42 AM EDTConviction: 100 / 100
Source: John Schulman @johnschulman2 Feb 7, 2025 Confirming that I left Anthropic last week. Leaving wasn't easy because I en...
GOOGrightbacktest DEMOTE

John Schulman confirms he left Anthropic (private) for another, unnamed opportunity. No details on destination or scope, so market impact on public equities is indirect and low-signal.

Mentioned: Jun 18, 2026, 12:42 AM EDTConviction: 16 / 100Return: -7.30%
Source: John Schulman @johnschulman2 Feb 7, 2025 Confirming that I left Anthropic last week. Leaving wasn't easy because I en...
AMZNrightbacktest DEMOTE

John Schulman confirms he left Anthropic (private) for another, unnamed opportunity. No details on destination or scope, so market impact on public equities is indirect and low-signal.

Mentioned: Jun 18, 2026, 12:42 AM EDTConviction: 18 / 100Return: -7.21%
Source: John Schulman @johnschulman2 Feb 7, 2025 Confirming that I left Anthropic last week. Leaving wasn't easy because I en...
OPENAIopen

John Schulman @johnschulman2 Aug 6, 2024 I shared the following note with my OpenAI colleagues today: I've made the d... John Schulman @johnschulman2 Aug 6, 2024 I shared the following note with my OpenAI colleagues today: I've made the difficult decision to leave OpenAI. This choice stems from my desire to deepen my focus on AI...

Mentioned: Jun 18, 2026, 12:41 AM EDTConviction: 100 / 100
Source: John Schulman @johnschulman2 Aug 6, 2024 I shared the following note with my OpenAI colleagues today: I've made the d...
AMZNrightbacktest PROMOTE

John Schulman announces he is leaving OpenAI to focus more on AI alignment and hands-on technical work. This is a high-profile personnel move at a private company; direct tradable impact is limited and second-order (talent/strategy perception around AI safety).

Mentioned: Jun 18, 2026, 12:41 AM EDTConviction: 35 / 100Return: +15.68%
Source: John Schulman @johnschulman2 Aug 6, 2024 I shared the following note with my OpenAI colleagues today: I've made the d...
GOOGLrightbacktest PROMOTE

John Schulman announces he is leaving OpenAI to focus more on AI alignment and hands-on technical work. This is a high-profile personnel move at a private company; direct tradable impact is limited and second-order (talent/strategy perception around AI safety).

Mentioned: Jun 18, 2026, 12:41 AM EDTConviction: 38 / 100Return: +14.93%
Source: John Schulman @johnschulman2 Aug 6, 2024 I shared the following note with my OpenAI colleagues today: I've made the d...
NVDArightbacktest PROMOTE

John Schulman announces he is leaving OpenAI to focus more on AI alignment and hands-on technical work. This is a high-profile personnel move at a private company; direct tradable impact is limited and second-order (talent/strategy perception around AI safety).

Mentioned: Jun 18, 2026, 12:41 AM EDTConviction: 43 / 100Return: +11.17%
Source: John Schulman @johnschulman2 Aug 6, 2024 I shared the following note with my OpenAI colleagues today: I've made the d...
MSFTrightbacktest DEMOTE

John Schulman announces he is leaving OpenAI to focus more on AI alignment and hands-on technical work. This is a high-profile personnel move at a private company; direct tradable impact is limited and second-order (talent/strategy perception around AI safety).

Mentioned: Jun 18, 2026, 12:41 AM EDTConviction: 39 / 100Return: -1.55%
Source: John Schulman @johnschulman2 Aug 6, 2024 I shared the following note with my OpenAI colleagues today: I've made the d...
MSFTrightbacktest DEMOTE

John Schulman announces he is leaving OpenAI to focus more on AI alignment and hands-on technical work. This is a high-profile personnel move at a private company; direct tradable impact is limited and second-order (talent/strategy perception around AI safety).

Mentioned: Jun 18, 2026, 12:41 AM EDTConviction: 39 / 100Return: +1.55%
Source: John Schulman @johnschulman2 Aug 6, 2024 I shared the following note with my OpenAI colleagues today: I've made the d...
NVDArightbacktest PROMOTE

Post announces the author is advising an unnamed company/product (link only) that uses AI to help scientific authors/reviewers perform deeper analysis than humans unaided. It’s a positive qualitative endorsement of AI in scientific publishing/workflows, but contains no financial details, adoption metrics, customer names, pricing, or public-company identifiers.

Mentioned: May 28, 2026, 11:46 AM EDTConviction: 24 / 100Return: +6.92%Observed price: $213.98
Source: Glad to be advising https://t.co/QVsH1u528N, which uses AI to help authors and reviewers do deeper, more thorough ana...
GOOGLrightbacktest PROMOTE

Post announces the author is advising an unnamed company/product (link only) that uses AI to help scientific authors/reviewers perform deeper analysis than humans unaided. It’s a positive qualitative endorsement of AI in scientific publishing/workflows, but contains no financial details, adoption metrics, customer names, pricing, or public-company identifiers.

Mentioned: May 28, 2026, 11:46 AM EDTConviction: 26 / 100Return: +7.88%Observed price: $390.86
Source: Glad to be advising https://t.co/QVsH1u528N, which uses AI to help authors and reviewers do deeper, more thorough ana...

About this channel

johnschulman2 posts analysis and commentary on AI research and product implications. Writing emphasizes how improvements in foundation models and real‑time multimodal interfaces drive demand for training/inference compute, hosting, tooling, and distribution platforms. Public posts include qualitative endorsements but generally lack granular commercial metrics.

Subscribersn/a
Videosn/a
Win rate82%
Average return+2.39%

@johnschulman2

Unlock the full track record

Follow @johnschulman2 for ongoing observations about AI model capabilities, multimodal interaction, and infrastructure implications.

2 more thesis calls are available after sign-up.