Google AI pioneer Le Viet Quoc leaves to co-found scientific research startup Discovery Loop

Four of Google's top AI researchers left to launch Discovery Loop, a startup building automated systems to speed up scientific research. The venture, announced August 5, secured funding from Alphabet.

Categorized in: AI News Science and Research
Published on: Aug 10, 2026
Google AI pioneer Le Viet Quoc leaves to co-found scientific research startup Discovery Loop

Four of Google's most prominent AI researchers have left the company to launch Discovery Loop, a startup building automated systems to accelerate scientific and engineering research. The venture, announced on August 5, has secured funding from Alphabet, Google's parent company, according to TechCrunch.

The founding team consists of Le Viet Quoc, Jeff Dean, Sanjay Ghemawat, and Oriol Vinyals. Jeff Dean, who joined Google in 1999 as its 30th employee and helped build core infrastructure for Google Search, will serve as CEO, according to sources familiar with the matter. Le Dean later led Google's early AI research efforts and played a central role in developing its multimodal Gemini models.

Le Viet Quoc earned his doctorate in computer science from Stanford and was named one of the world's top innovators under 35 by MIT Technology Review. He previously held a distinguished research position at Google.

What Discovery Loop plans to build

The startup's stated goal is to automate machine learning, engineering, and scientific research. At the core of their approach is solving what the team calls the "experimental loop" - the cycle of proposing experiments, running them, analyzing results, and refining methods.

The team described current manual experimental workflows as "agonizingly slow and labor-intensive." To address this, Discovery Loop is deploying an automated platform that uses frontier AI models and large-scale computing infrastructure to run thousands of experiments in parallel. The goal is to shorten iteration times while improving both quantity and quality of research output.

For Research professionals, this approach directly addresses a persistent bottleneck: the slow cycle of hypothesis testing and methodology refinement that constrains progress in many scientific fields.

A wave of departures from Google

Le Dean has faced a wave of high-profile exits from its research division in recent months. In June, prominent researchers Noam Shazeer and Jon Jumper resigned within days of each other to join competing AI firms, including OpenAI and Anthropic.

Google CEO Sundar Pichai publicly acknowledged Dean's departure in an August 6 blog post, calling it an "incredible 27-year journey" and saying the company was "excited to support his next chapter."

Why this matters for Science and Research jobs

This venture signals that senior AI researchers see scientific automation as the next major commercial opportunity in AI. For professionals currently running manual experimental workflows, the systems Discovery Loop is building could alter how research cycles operate. Monotonous, labor-intensive iterations that once took weeks could happen in parallel. This shifts the skill demand from executing experiments toward designing automated pipelines and interpreting results at a larger scale. The founding team includes three of the most cited AI researchers globally and two of the most cited figures in distributed computing, which suggests the technological capability to deliver on this premise is real.


Get Daily AI News

Your membership also unlocks:

700+ AI Courses
700+ Certifications
Personalized AI Learning Plan
6500+ AI Tools (no Ads)
Daily AI News by job industry (no Ads)