← The Independent 100: a study

Jürgen Schmidhuber

#2 in The Canon, the curator’s rank and section @SchmidhuberAI Relevance to our work: 3 of 5

Bracketed tags are evidence labels: [Verified] means we read the page, file or API response ourselves, [Reported] means a source says so and we did not check the underlying thing, and [Inference] is our own reasoning.

Credit

The Independent 100 list was curated by dan (@irl_danB) and published on 2026-09-24 at independent.prose.md (announcement). The ranks and section names are his, and number 100 is his open wild-card slot. This study is ours and is not endorsed by him or by anyone on the list.

What they publicly do

AI researcher who writes an AI blog of historical and technical notes, many issued as IDSIA technical notes. In it he attributes LSTM, Highway Nets, fast-weight programmers, adversarial curiosity, world models and early recursive self-improvement algorithms to work by his teams [Verified: people.idsia.ch/~juergen/blog.html]. The blog lists him as Director of the AI Initiative at KAUST from 2021 and as Chief Scientist of NNAISENSE [Verified: same page].

Public work

  • Recursive Self-Improvement (RSI) since 1987 (Technical Note IDSIA-9-26) paper CC BY-NC-SA 4.0 (blog) last activity 2026-09

    A September 2026 survey of RSI lines of work: meta-evolution, self-modifying policies, the Gödel Machine and curiosity-driven RSI [Verified: blog summary].

  • Annotated History of Modern AI and Deep Learning paper no license stated last activity 2025-12-29

    A survey that credits early contributors to deep learning, updated 2025-12-29 [Verified: blog entry].

  • AI Blog blog CC BY-NC-SA 4.0 last activity 2026-09

    Priority and history notes ('Who invented JEPA?', 'Who invented Transformer neural networks?', and the world-model boom). The blog states a CC BY-NC-SA 4.0 license [Verified].

  • EvoTorch (NNAISENSE) repo Apache-2.0 last activity 2026-08-31 · 1,148 GitHub stars on 2026-09-24

    A GPU-accelerated library of evolutionary algorithms built on PyTorch and Ray. The blog credits its release to NNAISENSE, where he is Chief Scientist [Verified: blog, GitHub API].

Licenses of GitHub repositories are from the authenticated GitHub license API on 2026-09-25, one row per repository in evidence/licenses.jsonl (“no license file” means the API found none; “license file present, not recognised by GitHub” means GitHub could not classify it). Other licenses are as the source states. Check the license file before reusing anything.

Relevance to our work

3 of 5 lenses: research, research-method

The RSI survey and the Gödel Machine lineage give the theory behind self-improving agent loops. His 'who invented X' notes are a strong example of prior-art tracing. EvoTorch is open under Apache-2.0 but falls outside our current lanes.

Scored 0 to 5 against the study’s relevance rubric: 5 means public, openly licensed work squarely in our lanes that we could study or adopt now, and 1 means little connection to our work. The score is one reviewer’s judgement of how much this public work bears on our own work. It is not a ranking of people and says nothing about the person or the quality of their work.

What we would study next

  1. Read the RSI technical note, focusing on the Gödel Machine's 'provably useful self-rewrite' criterion (people.idsia.ch/~juergen/recursive-self-improvement.html). It gives a formal bar for when a self-modification should be accepted. That bar applies to any loop in which agents rewrite their own skills or prompts. Effort: medium.
  2. Read his 'who invented X' notes as a template for a prior-art check (people.idsia.ch/~juergen/who-invented-transformer-neural-networks.html). The notes trace claims back to their dated primary sources, which is the check our rulebook wants before a [Verified] novelty claim. They argue his side of disputed priority claims, so treat them as one input, not a verdict [Inference]. Effort: small.

Before reusing anything: The blog is CC BY-NC-SA 4.0: non-commercial, share-alike. Its text cannot be reused in commercial content. EvoTorch is Apache-2.0.

Sources

Overall confidence in this record: High. Sources were read on 2026-09-24 and every link was checked on 2026-09-25. The full record is line 2 of people.jsonl.

Limits

Not verified: Whether he maintains any personal GitHub code. The RSI note itself was not read in full; only the blog summary was.

This record comes from a first-pass survey of public material by an AI agent session on 2026-09-24, edited for publication on 2026-09-25. Nothing was installed or run for it. Roles, stars and licenses change; see the study’s limits.

Request a correction

Is something about this record wrong or out of date, or would you rather not be covered? Open a correction (a GitHub issue form with this page’s address filled in). A request to remove a record is honoured without argument; accepted corrections are recorded in the changelog with the date.

Help fix it

Each opens a GitHub issue form with this page’s address filled in. You need a GitHub account.

Made by Joshua Nowak (ZestStream) with AI coding agents. Not affiliated with or funded by Jeffrey Emanuel; Joshua uses his public tools and is a paying subscriber to jeffreys-skills.md.

Source on GitHub MIT licensed