# Three steps for OCaml to crest the AI humps

*2025-10-01 — paper*

Authors: Sadiq Jaffer, Jon Ludlam, Ryan T. Gibb, Thomas Gazagnaire, Anil Madhavapeddy

Links: [PDF](https://anil.recoil.org/papers/2025-ocaml-ai.pdf), [BIB](https://anil.recoil.org/papers/2025-ocaml-ai.bib), [URL](https://conf.researchr.org/details/icfp-splash-2025/ocaml-2025-papers/5/Three-steps-for-OCaml-to-crest-the-AI-humps) (conf.researchr.org)

## Abstract


We discuss how OCaml could adapt to the fast-moving world of AI-assisted
agentic coding.

We first benchmark how well represented OCaml is in the large
and diverse set of open weight models that can be run locally. We then consider
what is unique about OCaml programming (in particular, modules and abstraction)
that differentiates it in this space. We then consider the changes required in
our ecosystem to work better with AI coding assistants.

See also [Sadiq's notes](https://toao.com/blog/ai-existential-ocaml) on the paper.


Classification: short
Venue: proceedings of the 2025 OCaml Workshop
URL: https://conf.researchr.org/details/icfp-splash-2025/ocaml-2025-papers/5/Three-steps-for-OCaml-to-crest-the-AI-humps
Projects: OCaml Labs, OxCaml Labs

Discussion:
- Twitter: <https://x.com/avsm/status/2019486202898899023>

## Related

- [A Decade of Docker Containers on the CACM cover!](https://anil.recoil.org/notes/cacm-docker-cover) (note, 2026-02-24)
  Our CACM cover article reflects on a decade of Docker, from the early days of hacking Docker for Mac on a French farm to today's AI-driven sandboxing, covering the technical origins, cross-platform challenges, and the vibrant open-source community that made it all possible.
- [2025 Advent of Agentic Humps: Building a useful O(x)Caml library every day](https://anil.recoil.org/notes/aoah-2025) (note, 2025-12-26)
  An exploration of agentic programming through building useful OCaml libraries daily using Claude Code while establishing groundrules for responsible development.
- [Three steps for OCaml to crest the AI humps](https://anil.recoil.org/videos/59166959-f1b2-431b-ae53-c126664e2346) (video, 2025-12-05)
  Three steps for OCaml to crest the AI humps, OCaml 2025 by Sadiq Jaffer, Jonathan Ludlam, Ryan Gibb, Thomas Gazagnaire, and Anil Madhavapeddy. We discuss how OCaml could adapt to the fast-moving world of AI-assisted agentic coding. We first benchmark how well represented OCaml is in the large and diverse set of open weight models that can be run locally. We then consider what is unique about OCaml programming (in particular, modules and abstraction) that differentiates it in this space. We then consider the changes required in our ecosystem to work better with AI coding assistants. Presentation at the OCaml 2025 workshop, Oct 17, 2025, https://conf.researchr.org/home/icfp-splash-2025/ocaml-2025.
- [Three Steps for OCaml to Crest the AI Humps](https://www.toao.com/blog/ai-existential-ocaml) (feed, 2025-10-24)
  Sadiq Jaffer. OCaml represents less than 0.01% of training data for many large language models, creating an existential threat for the language community. This talk outlines three concrete steps the OCaml community can take to improve AI coding agent support and ensure the language's future relevance.
- [What I learnt at ICFP/SPLASH 2025 about OCaml, Hazel and FP](https://anil.recoil.org/notes/icfp25-what-i-learnt) (note, 2025-10-09)
  Highlights from ICFP/SPLASH 2025 including Hazel live programming, OCaml AI tooling, formally verified GC, and cross-community discussions between Haskell and OCaml.
- [Three steps for OCaml to crest the AI humps](https://ryan.freumh.org/papers/2025-ocaml-ai.html) (feed, 2025-10-01)
  Ryan Gibb. Oct. 2025Sadiq Jaffer, Jon Ludlam, Ryan Gibb, Thomas Gazagnaire, and Anil Madhavapeddy.Proceedings of the 2025 OCaml workshop. We discuss how OCaml could adapt to the fast-moving world of AI-assisted agentic coding. We first benchmark how well represented OCaml is in the large and diverse set of ope…
- [OxCaml Labs](https://anil.recoil.org/projects/oxcaml) (project, 2025-01-01)
- [OCaml Labs](https://anil.recoil.org/projects/ocamllabs) (project, 2012-01-01)

---
Canonical: https://anil.recoil.org/papers/2025-ocaml-ai
Type: paper
License: CC BY 4.0 <https://creativecommons.org/licenses/by/4.0/>
Tags: ocaml, llms, ai, conference, ocamllabs, oxcaml