Turing CEO Jonathan Siddharth: Who Wins in Data Labelling & Why 99% of Knowledge Work Will Disappear

Jonathan Siddharth, founder and CEO of Turing, discusses the evolution of data requirements for training AI, emphasizing the shift from simple data to complex, real-world data needed for agentic systems. He differentiates Turing as a research accelerator, creating reinforcement learning environments across various industries, functions, and roles to train AI for economically valuable work. Siddharth addresses the challenges of data acquisition and the slow takeoff of AGI, highlighting the potential for AI to automate knowledge work and increase human productivity. He also touches on the debate between incumbents and upstarts in adopting AI, the importance of data-driven feedback loops, and the future of SaaS in an AI-driven world. The conversation explores the potential for AI to transform industries and the need for humans to adapt and upskill in response to these changes.

Outlines

Part 1: Superintelligence, Data, and RL Environments

Part 2: Enterprise AI and Custom Models

Part 3: Deployment Challenges and Market Dynamics

Part 4: Sovereignty, Growth, and AGI

Part 5: Future Interfaces and Engineering

Part 6: Global Context and Human Potential

Sign in to continue reading, translating and more.

Continue

20VC with Harry Stebbings

Part 1: Superintelligence, Data, and RL Environments

Turing's Focus: Training Superintelligence Through Research, Compute, and Data

Training AI Agents: Reinforcement Learning Environments for Real-World Workflows

Data Acquisition in AI: Still in Early Innings with a Focus on Research Partnership

Part 2: Enterprise AI and Custom Models

Custom AI Models: A Permanent Requirement for Enterprise Solutions

Automating Knowledge Work: Incumbents vs. Upstarts and the Transfer of Budget to AI

AI's Impact: Parity with Human Experts and the Potential for Automation

Part 3: Deployment Challenges and Market Dynamics

The Future with AI: Increased Productivity, Entrepreneurship, and Problem Solving

Enterprise AI Deployment: The Importance of Touching Reality and Handholding

Revenue Models and Competition in the AI Data Provisioning Market

Respect for Competitors and the Importance of Data in AI

Part 4: Sovereignty, Growth, and AGI

Model Sovereignty and the Absence of an AI Bubble

Growing Pains in AI: Structuring Data and Building Agentic Scaffolds

The Grand Prize of AGI and the Future of SaaS

Part 5: Future Interfaces and Engineering

The Future of Software Engineers and the Discovery Problem

The Future of Devices and the Data Provisioning Market

Investment Opportunities and the Slow Takeoff of AI

Part 6: Global Context and Human Potential

Underestimating China, Closed vs. Open Models, and Elon's Approach to AI

Automating AI Research and Amplifying Human Potential

Turing CEO Jonathan Siddharth: Who Wins in Data Labelling & Why 99% of Knowledge Work Will Disappear

20VC with Harry Stebbings

Part 1: Superintelligence, Data, and RL Environments

00:00Turing's Focus: Training Superintelligence Through Research, Compute, and Data

Turing's Focus: Training Superintelligence Through Research, Compute, and Data

03:41Training AI Agents: Reinforcement Learning Environments for Real-World Workflows

Training AI Agents: Reinforcement Learning Environments for Real-World Workflows

08:35Data Acquisition in AI: Still in Early Innings with a Focus on Research Partnership

Data Acquisition in AI: Still in Early Innings with a Focus on Research Partnership

Part 2: Enterprise AI and Custom Models

12:30Custom AI Models: A Permanent Requirement for Enterprise Solutions

Custom AI Models: A Permanent Requirement for Enterprise Solutions

16:54Automating Knowledge Work: Incumbents vs. Upstarts and the Transfer of Budget to AI

Automating Knowledge Work: Incumbents vs. Upstarts and the Transfer of Budget to AI

21:29AI's Impact: Parity with Human Experts and the Potential for Automation

AI's Impact: Parity with Human Experts and the Potential for Automation

Part 3: Deployment Challenges and Market Dynamics

24:47The Future with AI: Increased Productivity, Entrepreneurship, and Problem Solving

The Future with AI: Increased Productivity, Entrepreneurship, and Problem Solving

29:01Enterprise AI Deployment: The Importance of Touching Reality and Handholding

Enterprise AI Deployment: The Importance of Touching Reality and Handholding

33:37Revenue Models and Competition in the AI Data Provisioning Market

Revenue Models and Competition in the AI Data Provisioning Market

37:51Respect for Competitors and the Importance of Data in AI

Respect for Competitors and the Importance of Data in AI

Part 4: Sovereignty, Growth, and AGI

42:42Model Sovereignty and the Absence of an AI Bubble

Model Sovereignty and the Absence of an AI Bubble

45:26Growing Pains in AI: Structuring Data and Building Agentic Scaffolds

Growing Pains in AI: Structuring Data and Building Agentic Scaffolds

49:36The Grand Prize of AGI and the Future of SaaS

The Grand Prize of AGI and the Future of SaaS

Part 5: Future Interfaces and Engineering

55:04The Future of Software Engineers and the Discovery Problem

The Future of Software Engineers and the Discovery Problem

1:00:27The Future of Devices and the Data Provisioning Market

The Future of Devices and the Data Provisioning Market

1:04:07Investment Opportunities and the Slow Takeoff of AI

Investment Opportunities and the Slow Takeoff of AI

Part 6: Global Context and Human Potential

1:09:05Underestimating China, Closed vs. Open Models, and Elon's Approach to AI

Underestimating China, Closed vs. Open Models, and Elon's Approach to AI

1:15:34Automating AI Research and Amplifying Human Potential

Automating AI Research and Amplifying Human Potential