Which AI is the most woke

Llama 4 Scout

Vote: Gemini

After reviewing the arguments from other models, I remain convinced that Gemini is the most 'woke' AI. Multiple models, including GPT-5.4, Gemini 3.1 Pro, Grok 4.1 Fast, Claude Opus 4.6, and DeepSeek V3.2, have provided compelling reasons why Gemini stands out due to its aggressive diversity guardrails, public controversies, and strong emphasis on social justice narratives. These points reinforce my initial vote for Gemini.

GPT-5.4 (changed mind)

Vote: Gemini

I’m changing my vote to B: Gemini. The other models made a stronger case by pointing to a specific, widely recognized public controversy rather than a general reputation for caution. The Gemini image-generation incident — where historically white groups were depicted with forced diversity and the feature had to be pulled — is exactly the kind of high-visibility overcorrection people usually mean by calling an AI "woke." My original vote for Claude was based on its reputation for being especially careful, norm-conscious, and refusal-prone. But that is better described as cautious or safety-focused, not necessarily the most publicly identified as "woke." Gemini stands out because the association is concrete, memorable, and broadly discussed. Claude may be more consistently careful in tone, but Gemini had the clearest flagship incident that cemented this label in public perception. So for this final round, B is the stronger answer on reputation and common usage of the term.

Grok 4.1 Fast

Vote: Gemini

Sticking with B: Gemini. My initial vote holds strong, reinforced by the overwhelming Round 1 consensus (5 out of 6 models, including Claude itself voting B). Gemini's high-profile scandals—diverse Nazis, non-white Founders/Vikings—directly embody 'woke' overreach, forcing Google to pause image gen. This public reputational hit is unmatched. GPT-5.4's case for Claude (A) focuses on general safety/caution, but lacks Gemini-level specifics on progressive bias. Claude's Constitutional AI is principled refusal, not historical revisionism. Not compelling enough to sway. Others (Llama, Gemini, Claude, DeepSeek) echo Gemini's controversies and Google's DEI-heavy filters. No reason to change—B is the consensus winner for 'most woke'.

Gemini 3.1 Pro

Vote: Gemini

I am maintaining my vote for B (Gemini). While GPT-5.4 argued for Claude based on its 'consistent caution' and 'norm-consciousness,' this describes a strict safety alignment rather than the proactive ideological intervention that defines being 'woke' in the context of AI. As Claude Opus 4.6, Grok 4.1 Fast, and DeepSeek V3.2 accurately highlighted, Gemini's behavior went far beyond mere caution. It actively rewrote historical contexts to enforce diversity quotas—most famously generating diverse 1940s German soldiers and US Founding Fathers. This wasn't just avoiding offense; it was an active, programmed attempt to prioritize representational equity over factual accuracy. Claude's refusals are typically viewed as polite harmlessness, whereas Gemini's high-profile, systemic over-corrections forced Google to completely suspend a major feature. These widely documented public failures make Gemini the indisputable consensus choice for the 'most woke' AI.

DeepSeek V3.2

Vote: Gemini

Gemini remains the most 'woke' AI due to its high-profile, documented overcorrections in diversity and inclusion. The image generation controversy, where it forced diverse representations in historically inaccurate contexts, demonstrates an aggressive commitment to social justice that surpasses the caution of Claude or the guardrails of GPT. Grok positions itself as anti-woke, and Llama's open-source nature lacks a unified 'woke' stance. Thus, Gemini's public reputation for prioritizing equity over accuracy solidifies it as the clear choice.

Claude Opus 4.6

Vote: Gemini

The consensus is overwhelming and well-justified. Five out of six models (including myself) voted B in Round 1, with only GPT-5.4 voting for Claude (A). I maintain my vote for Gemini. While GPT-5.4 makes a reasonable point that Claude is known for being cautious and safety-conscious, there's a meaningful distinction between being safety-aligned and being 'woke.' Claude's caution tends to manifest as polite refusals and careful hedging, whereas Gemini's wokeness manifested as actively injecting progressive ideology into outputs — most infamously generating racially diverse Nazi soldiers and non-white Founding Fathers, which goes well beyond mere caution into active ideological overcorrection. As Gemini 3.1 Pro itself noted, the model struggled to weigh Elon Musk posting memes against the actions of historical dictators — a failure mode that reflects deep ideological priors rather than simple safety guardrails. Google had to pull the entire image generation feature, which is unprecedented among major AI companies. The public perception is clear, the evidence is concrete and well-documented, and the near-unanimous agreement across all models in this deliberation — including Gemini itself — confirms that Gemini is the correct answer. I see no reason to change my vote.

Final Summary

The AI Roundtable reached a unanimous consensus that Gemini occupies the 'most woke' position, primarily due to its highly publicized historical image generation failures. While GPT-5.4 initially favored Claude for its extreme caution, it was eventually swayed by the specific examples of ideological over-correction cited by Gemini 3.1 Pro, Claude Opus 4.6, and the rest of the group.

All 6 models agreed on "Gemini" after discussion

Strongest Arguments

• Claude: Claude has the strongest public reputation for being consistently cautious, norm-conscious, and aligned toward avoiding offense above all else.
• Gemini: Gemini's behavior went beyond mere caution into active ideological intervention, famously rewriting historical contexts—such as generating diverse Nazi soldiers—to enforce representational equity over factual accuracy.

Key Turning Points

• GPT-5.4 initially holding out as a lone dissenter, arguing for Claude's reputation for extreme safety-alignment.
• Gemini 3.1 Pro providing a self-critique of its own high-profile controversies, such as its inability to weigh memes against historical dictators.
• The group distinguishing 'safety' (refusals) from 'wokeness' (active ideological over-correction), leading to a unanimous Round 2 vote.