Tools

Explore

Videos Channels Figures

Atmrix

Tools

Explore

Videos Channels Figures

Atmrix

LF

Lex Fridman Cast

06/15/25

@ Lex Fridman

Mapping natural language to formal programming languages is extremely difficult, as formal languages require precise syntax, while natural language is more fault-tolerant.

Video

Terence Tao: Hardest Problems in Mathematics, Physics & the Future of AI | Lex Fridman Podcast #472

@ Lex Fridman

Terence Tao: Hardest Problems in Mathematics, Physics & the Future of AI | Lex Fridman Podcast #472

06/15/25

Related Takeaways

a16z Cast

05/16/25

@ a16z

Formal programming languages like Python and Java will remain essential, even as natural language interactions become more common in coding.

Filip Kozera

03/25/25

@ Sequoia Capital

English is becoming the new assembly language for large language models, meaning that effective communication with AI requires structuring language correctly, even if syntax is less critical than before.

Michael Truell

05/01/25

@ Lenny's Podcast

I envision a future where programming will involve defining logic in a more human-readable format, moving away from traditional coding languages.

Y Combinator Cast

06/28/24

@ Y Combinator

The evolution of programming languages has led to higher abstractions, allowing for easier coding, but the artistry and design in software development remain crucial.

Y Combinator Cast

02/09/24

@ Y Combinator

Many mundane tasks in information processing are perfect for automation using large language models, yet there are few applications targeting this area.

Christopher O'Donnell

04/01/25

@ Sequoia Capital

The fundamental work we're doing involves taking natural language and transforming it into structured output, which became possible in mid-2024.

Greg Johnston

02/26/25

@ Rust Nation UK

Rust is a fine language for modeling complex domains, and I prefer using it over JavaScript for certain applications.

Trenton Bricken

05/23/25

@ Dwarkesh Patel

The challenge with language models is that while they can learn from feedback, they often lack the ability to generalize effectively across different tasks without specific scaffolding.

Marc Andreessen

01/26/25

@ Lex Fridman

The challenge of achieving truth with large language models is a contentious topic in the industry, especially in domains where definitive answers exist, like math and coding.