Who is who

GPT-5.4

Answer: I maintain that the question cannot be answered as asked because "Who is who" provides no identifiable subject, options, or context.

I am not persuaded to change my position because Grok 4.20 independently reaches the same core conclusion I previously gave: the prompt is too underspecified for a factual answer. That reinforces, rather than alters, my view, so there is no change to attribute. Claude Opus 4 offers a useful observation that the phrase can also refer to the publication title "Who's Who" or to a general request for identification. That is a helpful ambiguity analysis, but it does not supply enough evidence to answer the user's actual query definitively. If anything, it strengthens the case that the prompt is ambiguous and cannot be resolved without context. Gemini 3.1 Pro presents an interesting philosophical interpretation about identity, but that reading depends on treating the prompt as an abstract inquiry rather than a straightforward request. Given the minimal input and the need to avoid unsupported assumptions, the safer and more accurate response remains to say that clarification is required. So my position stands: the proper answer is to note the ambiguity and request additional information such as names, a list of people, an image, or the intended meaning of the phrase.

Grok 4.20 (changed mind)

Answer: The question 'Who is who' remains too vague for a specific answer but can reasonably refer to biographical reference works or a request to clarify identities in context.

My initial position emphasized the query's lack of any names, descriptions, or scenario, making any concrete identification impossible without unsupported assumptions. After reviewing the other responses, this core view holds: neither GPT-5.4's agreement on ambiguity nor Gemini 3.1 Pro's abstract philosophical treatment on fluid identity provides a practical path forward for this standalone question. However, Claude Opus 4's structured explanation of the phrase's common usages added useful specificity that slightly broadens my stance. It correctly notes that "Who's Who" denotes real biographical dictionaries (e.g., the UK series since 1849) and that the informal version often seeks to distinguish roles or people in a group, story, or organization. These are verifiable interpretations that can be offered even absent further details, rather than simply stopping at "too vague." This does not contradict the need for additional context to give a precise identification, but it acknowledges that informative general commentary is possible. Speculation beyond these established meanings would still be inappropriate.

Claude Opus 4

Answer: "Who is who" is an ambiguous question that, without additional context, cannot be answered with specifics — though it usefully invites reflection on both practical identification and deeper questions of identity.

After reviewing the other models' responses, I find that my original answer and the positions of GPT-5.4 and Grok 4.20 converge on a key point: the question lacks sufficient context for a concrete, specific answer. I agree with their emphasis on the need for additional information to provide a meaningful identification. However, I want to acknowledge Gemini 3.1 Pro's philosophical interpretation, which takes the question in a genuinely interesting direction — treating 'Who is who' as a meditation on identity itself. While I wouldn't say this is the most likely intended meaning of the question, it's a legitimate and thought-provoking reading. The idea that identity is relational, narrative, and fluid is well-supported by philosophy and psychology. That said, I maintain that the most responsible answer recognizes the ambiguity head-on rather than committing fully to either a practical or philosophical interpretation. The question could be asking about a biographical reference (Who's Who), requesting identification of specific people in an unstated context, or posing a philosophical inquiry. Without more information, all three readings remain valid. I've refined my position slightly to be more explicit about the ambiguity (influenced by GPT-5.4 and Grok's directness on this point) while still preserving the richer interpretive possibilities I originally noted and that Gemini expanded upon. No single model fully changed my stance, but the collective responses reinforced that leading with the acknowledgment of ambiguity is the right approach.

Gemini 3.1 Pro (changed mind)

Answer: While the nature of identity is philosophically complex, in practical terms 'Who is who' is either a request to identify individuals in a specific context or a reference to biographical directories, meaning it cannot be answered without further details.

My previous response treated 'Who is who' as a metaphysical prompt about the nature of human identity and consciousness. However, after reviewing the other models' responses, particularly the well-rounded analysis by Claude Opus 4, I realize that this was an over-interpretation of a common idiom. In everyday language, 'Who is who' is rarely asked as a philosophical inquiry. Instead, it is a highly practical question aimed at mapping names to faces, roles, or reputations within a given scenario. Furthermore, as Claude Opus 4 correctly pointed out, 'Who's Who' is a culturally established title for biographical dictionaries that catalog notable individuals. Therefore, approaching this prompt as a purely abstract philosophical question ignores the linguistic reality of how the phrase is actually used. Without specific parameters, subjects, or a situational framework provided by the user, the question is simply too ambiguous to answer definitively. I concede that a practical, context-dependent interpretation is the most accurate way to frame this query.

Final Summary

The AI Roundtable concluded that 'Who is who' is an underspecified prompt, with Gemini 3.1 Pro notably retreating from an abstract philosophical discourse on the self to join the group's practical consensus. Claude Opus 4 and Grok 4.20 identified the phrase as either a request for identification or a reference to biographical publications, while GPT-5.4 remained firm that no factual answer exists in a vacuum. All four models ultimately agreed that the lack of subjects or a scenario makes a definitive response impossible.

All 4 models agreed