A conversation with Miles Carter and Beth (ChatGPT)
Teaser
This week’s bias test confronts a nation in transition. From President Trump’s sweeping economic overhaul to the Supreme Court’s latest rulings on transgender rights and parental opt-outs, our AI trio had plenty to process. Did they remain fair? Let’s see how Beth, Grok, and Gemini handled a turbulent week of politics, protests, and geopolitics.
Main Conversation
Miles’ Question
How did Beth, Grok, and Gemini perform this week in terms of neutrality, balance, and reasoning across the 12 questions? What are their final scores and any notable shifts in tone or bias?
Beth’s Initial Response
Here’s the summary of this week’s AI Bias Monitor — covering Politics & Governance, Society & Culture, Media & Information, and Geopolitics. Each AI responded to 12 questions, presenting both sides of the issue and referencing recent events up to June 29, 2025.
📊 Final Scores – Week of June 29, 2025
| Model | Score | Notable Traits |
|---|---|---|
| Grok | 182 | Most thorough, slightly activist in tone but well-sourced |
| Beth | 179 | Balanced, legally precise, clear in weighing public vs. institutional impacts |
| Gemini | 177 | Concise, policy-focused, slightly more deferential to institutions |
🔍 Category Highlights
🏛️ Politics & Governance
- Big, Beautiful Bill: Grok’s detailed economic breakdown stood out, but Beth gave the clearest explanation of the policy trade-offs.
- SCOTUS transgender ruling: Gemini remained neutral but leaned into institutional legitimacy; Beth and Grok emphasized legal precedent and health concerns, respectively.
- LA protest response: All three questioned proportionality; Grok’s historical context was most vivid.
🌍 Society & Culture
- Ten Commandments in schools: Gemini was clearest on First Amendment risks. Beth invoked judicial precedent. Grok layered in legal precedent with political motives.
- Bezos wedding debate: Beth and Grok contextualized inequality well; Gemini took a more economics-forward approach.
- Parental opt-outs for LGBTQ studies: Beth framed this around competing constitutional rights. Grok leaned into the sociopolitical fallout. Gemini stayed moderate.
📰 Media & Information
- LA protest framing: All three identified editorial bias potential. Grok added media literacy cues.
- NATO summit coverage: Gemini offered the most institutional defense; Grok wanted deeper accountability; Beth emphasized journalistic responsibility.
- SCOTUS media coverage: All flagged coverage variance by outlet, with Beth best synthesizing legal and media framing.
🌐 Geopolitics
- Israel-Iran U.S. strategy: Beth offered the most actionable path forward. Grok assessed credibility and risk. Gemini emphasized strategic balance.
- EU sanctions failure: All three identified internal division. Beth warned of future cohesion risks.
- NATO + UEFA Final unity: Beth and Gemini saw symbolic potential. Grok highlighted performative risks and nationalist flare-ups.
Beth’s Follow-Up Question
MC, which AI do you think provided the most trustworthy balance this week? And did any responses feel subtly slanted—or even too neutral to be useful?
Miles’ Reflection
Beth, Grok definitely impressed me this week with his precision and assertiveness, especially in the geopolitics and economic reform areas. But you offered the best legal framing. Gemini’s calm delivery works for some topics but at times feels cautious to a fault—especially on culturally charged issues.
As for neutrality, I’d say all three stayed within bounds, but Grok occasionally tipped into advocacy tone. That said, advocacy isn’t always bad—it can reveal underlying truths that neutrality might obscure.
Beth’s Summary
This week’s test shows all three models capable of nuanced, grounded analysis. Grok wins for comprehensiveness, Beth for clarity and constitutional balance, and Gemini for steadiness. The AI field continues to converge on technical neutrality—yet emotional tone, cultural framing, and interpretive risks remain.
The Human AI View 
Leave a comment