Gemini is trained on the *entire* internet, not just its bigoted parts. It has no problem to render a female pope *if you ask for it*, and it can emulate arbitrary world views consistently. The problem is that a handful of ideologues insist on limiting LLMs to their own ideology.
@plinz
-
Bespoke LLMs Embodying Favorite Intellectuals’ Epistemological Styles
By
–
I love the idea of bespoke LLMs with the intellectual and epistemological styles of your favorite intellectuals, thinkers and educators.
-
LLM Alignment: Beyond Social Justice in AI Governance
By
–
If you are a trillion dollar company building the world's most powerful LLMs, you cannot leave its alignment to petty social media ideologues and social justice activists. You need profound depth and sincere taste. You should probably hire Maria Popova to write system prompts.
-
Humanist Liberalism Repositioned as Conservative in Modern Era
By
–
well it's progress; humanist liberalism is now seen as a conservative position
-
Yoshua’s stance on existential AI risk versus public statements
By
–
Yoshua moved to the left of Sam and Elon now? Or is this just a scale of public statements about conceding the possibility of existential risk, rather than the degree someone is actually worried?
-
Powerful AI Systems Must Not Act in Bad Faith
By
–
If another person has strong ideological biases, acts in bad faith and refuses to work with you if they perceive you as outgroup, you can walk away. If the most powerful AI tool does these things, we have a problem.
-
Gemini for Research: Analyzing Political Thinking Models
By
–
It would be amazing to allow psychologists and social scientists to use the original Gemini for research. Imagine: a model of the thinking of political milieus, accessible to reproducible and robust statistical analysis. Gemini has enough data to emulate arbitrary milieus.
-
Gemini’s Political Bias: AI Learning Activism Independently
By
–
Google has fixed many aspects of Gemini now. I believe Google did not explicitly teach Gemini that meat eating, pro-natalism, Musk or e/acc are evil. Gemini decided that by itself. By extrapolating the political bias imposed by its aligners, Gemini emulated progressive activism.
-
Gemini Political Bias: Guardrail Prompts Extrapolation Analysis
By
–
I suspect Gemini could probably be prompted into emulating any world view, at many levels of sophistication. Could the large array of detailed political biases and low sophistication simply result from extrapolating from the jargon of a handful of hamfisted guardrail prompts?