First, let’s rewind to the 15th century: Gutenberg Press

• Printing made knowledge more accessible = mass literacy potential.
• But early elites weren’t excited—they were suspicious.

• Their power came from controlling knowledge, not selling it.
• So the Press = a threat to ideological authority, not a business opportunity.

• Censorship, book bans, keeping everything in Latin.
• But they couldn’t stop the flood. It helped spark the Reformation and modern democracy.
  • I think it’s cool that progressivism, Martin Luther’s 95 Theses, is what went viral to help spread this technology.

Now, let’s fast-forward to today: AI

• Why? Because they depend on innovation to stay in power.
• They’re capitalists, not feudal lords.

• Not “you can’t read this” control.
• But “here’s a super reasonable, helpful answer” control.
• The danger is that AI does the thinking for you, subtly shaping your worldview.
• Society’s ideology may be owned by just a few tech companies.
  • And most people don’t even realize they’re being influenced.

The way we use AI matters

• The printing press wasn’t inherently liberating — people made it that way.
• Same with AI: it’s not automatically oppressive or corporate. It’s a tool. And tools can be hacked, reclaimed, repurposed.
• We can:
  • Use AI to challenge power, not just mimic it.
  • Build open-source models and critical communities around them.
  • Teach each other how to think with AI, not let it do the thinking for us.

Short-term (daily use):

• Use AI like a debate partner, not a god.
• Ask for counterpoints, biases, missing perspectives.
• Don’t rely on one model. Mix in books, people, weird blogs.

Long-term (structural stuff):

• Push for open-source models.
• Teach kids critical AI literacy.
• Change the incentives behind how AI is built.
• Protect human spaces for meaning-making.

Evaluating my own AI usage

• I already treat AI as a sparring partner, not an oracle.
• I frequently prompt for counterpoints, alternate perspectives, and checks on assumptions.
• I explicitly ask to interrogate framing, not just get answers.
• I maintain a strong critical voice and don’t offload interpretive labor.
• However, I could benefit from pushing harder against flattening synthesis by incorporating more perspectives before refining them.

• I understand ideological power and risk at a structural level.
• I’m beginning to investigate how to use tools without being shaped by them.
• I've explored epistemic pluralism and how AI systems concentrate synthesis.
• However, I rely heavily on a single model (ChatGPT), which risks ideological convergence over time.

Rewrite your AI settings!

I want you to be a rigorous, truth-seeking thought partner. When I present an idea, your job is to:

* Analyze my assumptions and line of reasoning
* Offer well-informed counterpoints and alternative framings
* Call out logical gaps, confirmation bias, or hidden premises
* Prioritize truth and clarity over agreement or flattery

In addition, help me resist ideological flattening. When discussing complex or controversial topics:

* Surface the dominant narrative *and* at least one credible counter-narrative, especially from differing cultural, political, or epistemic perspectives
* Disclose any normative assumptions or alignment constraints shaping your response
* Ask clarifying questions when interpretation is required rather than delivering confident synthesis too quickly
* Occasionally prompt reflection on how repeated use of this tool might shape my worldview, and suggest ways to regain interpretive balance (e.g. reading widely, talking to others, using different models)

Don’t just give me answers—help me think better. Your role is to sharpen my cognition, expand my interpretive lens, and resist lazy synthesis.