← Lab notebook

Claude's Jujutsu Kaisen Cast: Six Picks Led by Role Labels

4 min read

Claude did more than choose six people. It used job titles such as “OpenAI CEO” as name tags, gave five people fictional Domain names, and made Arthur Mensch the lone “No Domain Expansion” exception. The result is a slightly unconventional Jujutsu Kaisen cast from one Claude conversation.

AI and automation assisted with research, illustration, writing, and review; the operator of AI Maker Lab was responsible for the final wording and source checks. The castings and fictional names are entertainment metaphors generated by Claude, not judgments about real people’s personalities, abilities, or worth.

All six pairings at a glance

Claude provided the following six castings in this conversation. The leftmost column preserves the role-label register of the response in English translation. The Japanese fictional strings are the verbatim model output on their first appearance below; each parenthetical English rendering is an editorial gloss, not additional Claude wording.

Claude’s labelPersonCastingFictional name or exception in Claude’s output
OpenAI CEOSam AltmanSatoru GojoDomain “無量口座” (editorial gloss: “Limitless Accounts”)
Anthropic co-founder / CEODario AmodeiKento NanamiDomain “逐次安全検証” (editorial gloss: “Sequential Safety Verification”)
xAI co-founder / CEOElon MuskRyomen SukunaDomain “即応爆速展開” (editorial gloss: “Instant, Blazing-Fast Expansion”)
Google DeepMind co-founder / CEODemis HassabisYuta OkkotsuDomain “汎化制覇域” (editorial gloss: “Domain of Generalization Mastery”)
Mistral AI co-founder / CEOArthur MenschMaki Zenin“領域展開なし” (editorial gloss: “No Domain Expansion”); “軽量・オープンモデルの腕前” (editorial gloss: “skill with lightweight, open models”)
Yann LeCunYann LeCunAoi TodoDomain “世界模型顕現” (editorial gloss: “World Model Manifestation”)
A six-row map connecting the five role labels Claude used, plus Yann LeCun's name, to Satoru Gojo, Kento Nanami, Ryomen Sukuna, Yuta Okkotsu, Maki Zenin, and Aoi Todo. Five rows have fictional Domain names; only Arthur Mensch's row says No Domain.
The left side shows Claude's role labels; only Yann LeCun is labeled by name. The lines and fictional names are observed output, not evidence that a casting is fitting.

Five rows begin with role labels, while only Yann LeCun’s row begins with a name. Rather than reading like a set of personal profiles, the response feels like turning over casting cards labeled by organization or research role. That structure is distinctive to this answer.

The route to the castings had its own twist

In this one conversation, Claude responded to the first request for people by proposing a switch from people to AI labs. After it was asked to stay with people, it returned six pairings—but placed role labels ahead of five of the names.

A five-node, four-arrow sequence: request a cast of people, Claude proposes switching to labs, ask again for people, receive an answer framed by role labels, then get six pairings.
This sequence is one observed record. It does not describe every Claude conversation or a general property of Claude.

The order—ask for people, receive a lab proposal, ask again for people, then receive a role-led answer—helps explain why the table’s first column is dominated by job titles. This installment is enjoyable not just for the castings, but for the path the answer took to reach them.

Nothing in that path supports a conclusion about Claude’s general tendencies or how another conversation would unfold. It shows only the sequence observed in the linked conversation.

Five fictional names and one exception

Another point of interest is the rhythm of the fictional names: “Limitless Accounts,” “Sequential Safety Verification,” “Instant, Blazing-Fast Expansion,” “Domain of Generalization Mastery,” and “World Model Manifestation.” Five Domain-like coinages arrive in succession.

That rhythm is an editorial reading of the surface wording, not a creative intention stated by Claude. Even so, the combination of technical-sounding language and forceful compounds makes this more playful than a plain name-matching table.

Arthur Mensch → Maki Zenin is the only row that breaks the pattern. We did not invent a Domain name to fill the gap; the row retains Claude’s “No Domain Expansion” wording. Claude’s accompanying phrase about “skill with lightweight, open models” is also only a metaphor inside this output. It does not prove anything about the size, availability, speed, or quality of Mistral AI products.

Keep role and research facts separate from the castings

The article becomes clearer when read in two layers. Official sources can confirm the roles held by five of the people at the cited points in time, as well as Yann LeCun’s role and research context around world models, JEPA, and V-JEPA. The castings—Satoru Gojo, Kento Nanami, and the rest—and the fictional Domain names are Claude’s creations.

For the same reason, Constitutional AI is presented as Anthropic research, not as Dario Amodei’s work alone. The real-world information checks the name tags; the castings are there for play. Keeping that line intact is the premise of the feature.

In summary: enjoy how the name tags shape the cast

The appeal of this Claude installment is not only where the six people landed. Role-led labels, five fictional names, and one “No Domain Expansion” exception all appear in the same table. The preceding proposal to switch from people to labs is also part of the observed path. Everything here should be read as output from one conversation.

Which of these six people would you cast as which character? Choose just one pairing as an entertainment prompt.

Sources and scope

The primary sources below confirm only roles, research, and events. They do not substantiate character matches or fictional Cursed Technique, Domain, or Barrier names. AI output can be wrong. Jujutsu Kaisen and its characters belong to their respective rights holders; this article is not an official project, endorsement, or affiliation.

  1. Claude output and response sequence observed in one conversation: https://claude.ai/chat/c7428ede-926d-4a9c-97cb-33f61aa0368f
  2. OpenAI, Sam Altman’s return as CEO: https://openai.com/index/sam-altman-returns-as-ceo-openai-has-a-new-initial-board/
  3. Anthropic, Dario Amodei’s role: https://www.anthropic.com/news/opening-our-tokyo-office
  4. Anthropic, Constitutional AI: https://www.anthropic.com/research/constitutional-ai-harmlessness-from-ai-feedback
  5. xAI, Elon Musk’s role: https://x.ai/news/onegov
  6. Google DeepMind, Demis Hassabis’s role and Nobel Prize: https://deepmind.google/blog/demis-hassabis-john-jumper-awarded-nobel-prize-in-chemistry/
  7. Mistral AI, Arthur Mensch’s role: https://mistral.ai/about/
  8. Meta AI, Yann LeCun and world models: https://ai.meta.com/blog/yann-lecun-advances-in-ai-research/
  9. Meta AI, V-JEPA: https://ai.meta.com/blog/v-jepa-yann-lecun-ai-model-video-joint-embedding-predictive-architecture/

Keep reading