Choose boring technology and LLMs

The author expresses a preference for tried-and-tested technology, such as relational databases, SQL, HTTP, HTML, and well-known programming languages. They are particularly excited about the rise of Large Language Models (LLMs), which they believe are revolutionizing the tech world due to their ability to perform tasks previously thought impossible. The author identifies as a "prompt engineer," using LLMs to build digital products, and prefers to use simple, robust software like Go and SQLite on a VPS from Hetzner, only resorting to cloud technologies when necessary.

The author spends their time learning about integrating LLMs into applications, exploring areas like prompt engineering, evals, use of embeddings, and multi-modality of LLMs. They share some resources for learning about language models, including writings by Simon Willison, a book on prompt engineering for LLMs, and two free video courses: "Mastering LLMs" and "Neural Networks" by 3Blue1Brown.

Key takeaways:

The author appreciates tested and reliable technology like SQL, HTTP, HTML, and well-known programming languages, and is particularly interested in LLMs.
The author uses simple and robust software like Go and SQLite, and occasionally cloud technologies like S3/R2 for storing data.
The author is investing time in learning about integrating LLMs into applications, including prompt engineering, evals, use of embeddings, and multi-modality of LLMs.
The author recommends resources for learning about language models, including writings by Simon Willison, the book 'Prompt Engineering for LLMs', a free video course 'Mastering LLMs', and another free video course 'Neural Networks' by 3Blue1Brown.

Choose boring technology and LLMs

Key takeaways:

Comments (0)

Newsletter