Suggestions
Stanislas Polu
Software Engineer & Co-Founder, Dust
Stanislas Polu is a prominent software engineer and entrepreneur based in Paris, France. He is currently the Co-Founder and Software Engineer at Dust, a company he launched in September 2022. Prior to this, he spent over three years as a Research Engineer at OpenAI, where he contributed to advanced AI technologies from September 2019 to September 2022.12
Education
Polu holds a Master's degree in Computer Science from Stanford University (2007-2009) and a Bachelor's degree from École Polytechnique (2004-2007). He also participated in a summer research fellowship at Purdue University in 2006.1
Professional Experience
His career includes significant roles at various tech companies:
- Stripe: Software Engineer (February 2015 - July 2019) in San Francisco.
- TOTEMS: Co-Founder and Software Engineer (August 2010 - February 2015), focusing on data analytics.
- Oracle: Member of Technical Staff (July 2009 - June 2010).
- Apple: Software Engineer Intern (June 2008 - September 2008).
- Exalead: Software Engineer Intern (April 2007 - August 2007), where he worked on semantic extraction rules for HTML documents.
- Kleonet: Co-Founder (2004 - 2005), where he developed an online file storage service.1
Skills and Languages
Polu is proficient in both English and French, with full professional proficiency in English and native-level fluency in French. His technical expertise spans various programming languages and software development practices.1
Recent Activities
In recent months, Polu has been active on LinkedIn, sharing insights about his journey with Dust, including the rapid development of products and the dynamics of startup culture. He has expressed enthusiasm about the future of technology and its potential to disrupt traditional sectors, particularly finance.234
Highlights
This is such fantastic science of DL and human science at the same time.
The fact that there is scaling on pre 1931 data on coding evals is mind blowing.
If you think LLMs are just stochastic parrots representing probability distributions, that means that coding in TS was backed into human knowledge decades before it existed.
If you think that LLMs truly generalize to raw intelligence that’s an incredible confirmation.
But the interesting bit is that it’s one or the other and whichever it is, the conclusion is 🤯
