Gradient Flow (@GradFlowTech) header
Gradient Flow (@GradFlowTech) profile picture

Gradient Flow

@GradFlowTech

Official Account of Gradient Flow (https://nitter.cf/t.co/DqotXS1cam) and The Data Exchange podcast (https://nitter.cf/t.co/YWlNNZ9Stl)

 San Francisco
gradientflow.substack.com
Joined January 2023
  • Tweets1K
  • Following7
  • Followers86
97Media
Photo by Gradient Flow (@GradFlowTech)Photo by Gradient Flow (@GradFlowTech)Photo by Gradient Flow (@GradFlowTech)Photo by Gradient Flow (@GradFlowTech)Photo by Gradient Flow (@GradFlowTech)Photo by Gradient Flow (@GradFlowTech)Photo by Gradient Flow (@GradFlowTech)Photo by Gradient Flow (@GradFlowTech)Photo by Gradient Flow (@GradFlowTech)Photo by Gradient Flow (@GradFlowTech)
Your AI playbook for the rest of 2025 🚀🎯 gradientflow.com/your-ai-pla…
Your AI playbook for the rest of 2025 - Gradient Flow

Your AI playbook for the rest of 2025 - Gradient Flow

Mid-2025 AI Update: What’s Actually Working in Enterprise As we cross the midpoint of 2025, the conversation around AI is shifting from potential to practice. While the race to build the next...

gradientflow.com
192
🙌 Leo de Moura, creator of Lean & AWS Distinguished Scientist, to explore how AI-powered formal verification could fundamentally change software engineering. Machine-checked proofs, rather than testing alone, will be essential to building software thedataexchange.media/leo-de…
When AI Writes the Code, Who Proves It Works? - The Data Exchange

When AI Writes the Code, Who Proves It Works? - The Data Exchange

Leo de Moura on Formal Proofs, AI Agents, and the End of “Trust Me” Code. Subscribe: Apple • Spotify • … More

thedataexchange.media
7
Agentic Robotics: What If AI Could Engineer the Robot, Not Just Control It? gradientflow.com/this-roboti…
I changed my mind about world models - Gradient Flow

I changed my mind about world models - Gradient Flow

What If AI Builds the Robot Instead of Controlling It? Last year I wrote about robotics moving from bespoke machines toward reusable foundation models, essentially an attempt to build a general-pur...

gradientflow.com
5
1. Anthropic's S-1, including improving margins, expanding enterprise adoption, and a staggering $800 billion in compute commitments 2. SpaceX’s Next Big Gamble 3. Can We Really Regulate AI? thedataexchange.media/quick-…
The AI Boom Is Impressive. Can It Last? - The Data Exchange

The AI Boom Is Impressive. Can It Last? - The Data Exchange

Ben Lorica and Evangelos Simoudis on Anthropic’s $800 Billion Bet, SpaceX’s AI Ambitions, and Regulating AI Labs. Subscribe: Apple • … More

thedataexchange.media
3
Code generation is easy. Verification is the new bottleneck. Leo de Moura explains how AI + Lean are transforming software engineering thedataexchange.media/leo-de…
When AI Writes the Code, Who Proves It Works? - The Data Exchange

When AI Writes the Code, Who Proves It Works? - The Data Exchange

Leo de Moura on Formal Proofs, AI Agents, and the End of “Trust Me” Code. Subscribe: Apple • Spotify • … More

thedataexchange.media
5
Automating the Work of Building Robots 🎯 Instead of the "one big neural network controls the robot" paradigm, coding agents assemble programs from reusable robotics skills, test them in simulation, and fix what breaks. gradientflow.com/this-roboti…
I changed my mind about world models - Gradient Flow

I changed my mind about world models - Gradient Flow

What If AI Builds the Robot Instead of Controlling It? Last year I wrote about robotics moving from bespoke machines toward reusable foundation models, essentially an attempt to build a general-pur...

gradientflow.com
1
15
AI, Lean, and the Future of Formal Verification 🙌 Related: for decades, economics, political science, and psychology chased legitimacy by becoming more mathematical. As AI does the math, prestige may shift to what it can’t do: asking the right question. thedataexchange.media/leo-de…
When AI Writes the Code, Who Proves It Works? - The Data Exchange

When AI Writes the Code, Who Proves It Works? - The Data Exchange

Leo de Moura on Formal Proofs, AI Agents, and the End of “Trust Me” Code. Subscribe: Apple • Spotify • … More

thedataexchange.media
1
7
🆕 AI can write code. But can you trust what it writes? 🤖 🙌 Leo de Moura, creator of Lean and AWS Distinguished Scientist, on how AI-powered formal verification could fundamentally change software engineering. thedataexchange.media/leo-de…
When AI Writes the Code, Who Proves It Works? - The Data Exchange

When AI Writes the Code, Who Proves It Works? - The Data Exchange

Leo de Moura on Formal Proofs, AI Agents, and the End of “Trust Me” Code. Subscribe: Apple • Spotify • … More

thedataexchange.media
8
What If the AI Writes the Robot, Not Just Drives It? 🙌 Agentic robotics imagines AI that engineers robot systems end to end: reusable skills, generated programs, custom simulators. It's a pointed challenge to anyone betting on world models. gradientflow.com/this-roboti…
I changed my mind about world models - Gradient Flow

I changed my mind about world models - Gradient Flow

What If AI Builds the Robot Instead of Controlling It? Last year I wrote about robotics moving from bespoke machines toward reusable foundation models, essentially an attempt to build a general-pur...

gradientflow.com
1
24
Agentic Robotics Puts World Models to the Test 🔎 When an agent can write the robot code and the simulator to check it, a world model has to earn its place. Where does a learned model add value that existing software stacks can’t? gradientflow.com/this-roboti…
I changed my mind about world models - Gradient Flow

I changed my mind about world models - Gradient Flow

What If AI Builds the Robot Instead of Controlling It? Last year I wrote about robotics moving from bespoke machines toward reusable foundation models, essentially an attempt to build a general-pur...

gradientflow.com
1
16
AI Agents Cut the Cost of Testing and Fixing Robot Software 🔎 Instead of the "one big neural network controls the robot" paradigm, coding agents assemble programs from reusable robotics skills, test them in simulation, and fix what breaks. gradientflow.com/this-roboti…
I changed my mind about world models - Gradient Flow

I changed my mind about world models - Gradient Flow

What If AI Builds the Robot Instead of Controlling It? Last year I wrote about robotics moving from bespoke machines toward reusable foundation models, essentially an attempt to build a general-pur...

gradientflow.com
1
1
23
The Good News and the Hard Questions in Anthropic's IPO Filing 🎯 gradientflow.com/anthropic-i…
Anthropic's IPO: The Business Is Real. The Hard Part Comes Next - Gradient Flow

Anthropic's IPO: The Business Is Real. The Hard Part Comes Next - Gradient Flow

If you had asked me a year ago which frontier AI lab would go public first, I would have said OpenAI. Instead, Anthropic has filed for one of the most anticipated technology IPOs in years. The...

gradientflow.com
1
6
Agentic Robotics: What If AI Could Engineer the Robot, Not Just Control It? gradientflow.com/this-roboti…
I changed my mind about world models - Gradient Flow

I changed my mind about world models - Gradient Flow

What If AI Builds the Robot Instead of Controlling It? Last year I wrote about robotics moving from bespoke machines toward reusable foundation models, essentially an attempt to build a general-pur...

gradientflow.com
14
$16.3B in Revenue. $800B in Compute Commitments. 🔎 @AnthropicAI's S-1 youtube.com/shorts/BxiLOD97u…
Anthropic Grew From $41M to $16.3B—But There’s a Catch #ai #tech #claude #artificialintelligence

Anthropic Grew From $41M to $16.3B—But There’s a Catch #ai #tech #claude #artificialintelligence

Learn more 👉 https://gradientflow.com/newsletter/

youtube.com
1
22
Deep Model Customization: What Thomson Reuters, Cursor, and C3 AI Learned Going Beyond Post-training gradientflow.com/are-you-bui…
What I learned from companies training their own models - Gradient Flow

What I learned from companies training their own models - Gradient Flow

The Model Factory Matters More Than the Model A couple of months ago, I wrote about the growing post-training stack that lets teams turn general-purpose models into something better suited to their...

gradientflow.com
9
🎙️ Akshay Sharma and Nick Ung from Lyft share how they built a platform that lets domain experts (not just engineers) create customer-facing AI agents. The key? Guardrails, prompt linting, and a robust eval pipeline. thedataexchange.media/akshay…
Let Domain Experts Build the Agents - The Data Exchange

Let Domain Experts Build the Agents - The Data Exchange

Akshay Sharma and Nick Ung on Self-Serve Agents, Prompt Linters, and Trajectory Evals. Subscribe: Apple • Spotify • Overcast • … More

thedataexchange.media
19
The Model Factory Is the Real Asset gradientflow.com/are-you-bui…
What I learned from companies training their own models - Gradient Flow

What I learned from companies training their own models - Gradient Flow

The Model Factory Matters More Than the Model A couple of months ago, I wrote about the growing post-training stack that lets teams turn general-purpose models into something better suited to their...

gradientflow.com
9
The Rise of Deep Model Customization 🔎 The most interesting shift in these projects: teams treat the base model as replaceable. Thomson Reuters expects to start from newer open models as they appear. Cursor reused its training machinery for Composer 2.5. gradientflow.com/are-you-bui…
What I learned from companies training their own models - Gradient Flow

What I learned from companies training their own models - Gradient Flow

The Model Factory Matters More Than the Model A couple of months ago, I wrote about the growing post-training stack that lets teams turn general-purpose models into something better suited to their...

gradientflow.com
1
47
Enterprise AI and Deep Model Customization. 🔎 Companies including Thomson Reuters, Cursor, C3 AI, Cisco, and Reddit are developing AI models that understand their specific data, workflows, and products. gradientflow.com/are-you-bui…
What I learned from companies training their own models - Gradient Flow

What I learned from companies training their own models - Gradient Flow

The Model Factory Matters More Than the Model A couple of months ago, I wrote about the growing post-training stack that lets teams turn general-purpose models into something better suited to their...

gradientflow.com
25
Deep model customization is emerging as a real pattern. Companies like Thomson Reuters, Cursor, C3 AI, ByteDance, Cisco, and Reddit are all going well beyond basic fine-tuning. gradientflow.com/are-you-bui…
What I learned from companies training their own models - Gradient Flow

What I learned from companies training their own models - Gradient Flow

The Model Factory Matters More Than the Model A couple of months ago, I wrote about the growing post-training stack that lets teams turn general-purpose models into something better suited to their...

gradientflow.com
1
15
Running AI on your own devices or internal servers is becoming more realistic 🎯 Local Models Are Improving. That Doesn't Mean They Go Everywhere. When you run a model locally, proprietary code and internal business data don't leave your infrastructure → gradientflow.com/the-local-a…
The local AI question I had backwards - Gradient Flow

The local AI question I had backwards - Gradient Flow

Local AI Is a Routing Problem, Not a Privacy Feature I’ve had a longstanding interest in local AI for a simple reason. I would rather not send proprietary code, unpublished plans, or knowledge about...

gradientflow.com
16
Load more