Industry Newsreinforcement learningdata labelingturing
Turing CEO Declares Simple Labeling Obsolete
9.0
Jonathan Siddharth, CEO of Turing, said on the 20VC podcast Monday that basic data-labeling is becoming obsolete as modern AI models require complex, real-world training data and reinforcement-learning environments that mirror human knowledge work. He urged training firms to act as proactive research partners; Turing raised $111 million in June at a $2.2 billion valuation and reported a $300 million 2024 revenue run rate.
Key Points
- 1Rejects basic labeling: Turing CEO says simple annotation no longer meets modern model training needs
- 2Highlights complexity: Models require real-world tasks, reinforcement-learning environments, and domain experts for meaningful data
- 3Advises practitioners: Build simulated workflow environments and recruit subject-matter experts to accelerate research partnerships
Scoring Rationale
Strategic industry-level insight with high applicability, limited by reliance on executive commentary rather than independent validation
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
