/
Navigation
Chronicles
Browse all articles
Explore
Semantic exploration
Research
Entity momentum
Nexus
Correlations & relationships
Story Arc
Topic evolution
Drift Map
Semantic trajectory animation
Posts
Analysis & commentary
Pulse API
Tech news intelligence API
Browse
Entities
Companies, people, products, technologies
Domains
Browse by publication source
Handles
Browse by social media handle
Detection
Concept Search
Semantic similarity search
High Impact Stories
Top coverage by position
Sentiment Analysis
Positive/negative coverage
Anomaly Detection
Unusual coverage patterns
Analysis
Rivalry Report
Compare two entities head-to-head
Semantic Pivots
Narrative discontinuities
Crisis Response
Event recovery patterns
Connected
Search: /
Command: ⌘K
Embeddings: large
TEXXR

Chronicles

The story behind the story

days · browse · Enter similar · o open

OpenAI says it will migrate to Facebook's PyTorch machine learning framework for future projects, eschewing Google's TensorFlow platform

We are standardizing OpenAI's deep learning framework on PyTorch. Matt Asay / InfoWorld : Interested in machine learning? Better learn PyTorch Tweets: @openai : We're standardizing OpenAI's deep learning framework on PyTorch to increase our research productivity at scale on GPUs (and have just released a PyTorch version of Spinning Up in Deep RL): https://openai.com/... https://twitter.com/... Emil Protalinski / @epro : Big move: @OpenAI is ditching Google's @TensorFlow for Facebook's @PyTorch https://venturebeat.com/... Mike Schroepfer / @schrep : Looking forward to working even more closely together. Awesome to have you in the PyTorch community!! https://twitter.com/...

VentureBeat Kyle Wiggers

Context & Ripple Effects

OpenAI's move is the marquee confirmation of a shift already visible in the data: an analysis of AI research papers found TensorFlow dominant in industry while most researchers had already switched to PyTorch. Facebook spent 2018-2019 making that switch painless — PyTorch 1.0 fixed production performance and cross-framework compatibility, and tools like BoTorch and Ax extended the ecosystem into experiment management.

First-order effects

  • Google loses its highest-profile research lab as a TensorFlow showcase; Facebook gains a flagship user for PyTorch at exactly the moment Mike Schroepfer signals closer collaboration with OpenAI.

Second-order effects

  • Facebook doubles down on the ecosystem play — PyRobot for robotics and BoTorch/Ax for experiments give researchers fewer reasons to leave the framework — while Google must decide whether TensorFlow's industry base can survive losing the research community that feeds talent and paper mindshare.

Third-order effects

  • The pattern ends with PyTorch as neutral infrastructure rather than a Facebook asset: Meta later hands governance to the Linux Foundation's PyTorch Foundation alongside Google, AMD, Azure, and AWS, and a Microsoft-backed enterprise support program formalizes commercial adoption — framework competition settling into shared stewardship of one de facto standard.

The trend: ML frameworks are consolidating around PyTorch as the research-to-production standard, with vendors competing on ecosystems and governance rather than rival libraries.

Discussion

  • @openai @openai on x
    We're standardizing OpenAI's deep learning framework on PyTorch to increase our research productivity at scale on GPUs (and have just released a PyTorch version of Spinning Up in Deep RL): https://openai.com/... https://twitter.com/...
  • @epro Emil Protalinski on x
    Big move: @OpenAI is ditching Google's @TensorFlow for Facebook's @PyTorch https://venturebeat.com/...
  • @schrep Mike Schroepfer on x
    Looking forward to working even more closely together. Awesome to have you in the PyTorch community!! https://twitter.com/...