AI Papers to Read in 2025

This article offers reading suggestions to keep you updated on key breakthroughs in AI and Data Science, combining latest findings with classic research.

Returning to AI Paper Recommendations

After a long break, the author resumes their popular series of AI paper recommendations on Towards Data Science (TDS). Previous editions include four notable lists referenced as [1], [2], [3], and [4]. This renewed series aims to provide insightful perspectives rather than a simple catalog of state-of-the-art models.

Purpose and Approach

The list intends to encourage critical thinking about AI's current state. It consists of ten carefully selected papers, each accompanied by a brief outline of the paper's contributions and clear reasons for why it is valuable. Additionally, each paper features a further reading section covering relevant tangents for deeper exploration.

Author’s Perspective

“We don’t need larger models; we need solutions,” and “do not expect me to suggest GPT nonsense here.”

Originally stated in 2022, the author anticipated repeating this stance, believing that new GPT models would be bigger but only marginally better—not truly groundbreaking.

Nevertheless, the author acknowledges giving credit where it is due.

Summary

This curated list highlights meaningful AI research that offers real insights for what to watch in the upcoming years, helping readers stay informed without simply focusing on the newest massive models.

Towards Data Science — 2025-11-06