Menu
Sign In Search Podcasts Charts People & Topics Add Podcast API Blog Pricing

Narrator (TYPE III AUDIO)

👤 Speaker
266 total appearances

Appearances Over Time

Podcast Appearances

LessWrong (Curated & Popular)
“The Pando Problem: Rethinking AI Individuality” by Jan_Kulveit

These layers often align, but can conflict.

LessWrong (Curated & Popular)
“The Pando Problem: Rethinking AI Individuality” by Jan_Kulveit

In this ontology, the relationship between character and predictive substrate introduces more options of AI individuality, making things weirder still.

LessWrong (Curated & Popular)
“The Pando Problem: Rethinking AI Individuality” by Jan_Kulveit

One predictive ground, multiple characters.

LessWrong (Curated & Popular)
“The Pando Problem: Rethinking AI Individuality” by Jan_Kulveit

The same base AI model can be fine-tuned to produce distinctly different personas.

LessWrong (Curated & Popular)
“The Pando Problem: Rethinking AI Individuality” by Jan_Kulveit

This is clear with conversational instances understood as ephemeral activation patterns.

LessWrong (Curated & Popular)
“The Pando Problem: Rethinking AI Individuality” by Jan_Kulveit

If you change the system prompt to this is conversation between wise fairy tale squirrel and young hero, you get a different behavior than from this is conversation between helpful AI assistant and a user.

LessWrong (Curated & Popular)
“The Pando Problem: Rethinking AI Individuality” by Jan_Kulveit

But something similar is true also with deeper characters, like Claude Sonnet.

LessWrong (Curated & Popular)
“The Pando Problem: Rethinking AI Individuality” by Jan_Kulveit

the same weights would be able to support both a helpful polite AI character called Claude, and another character, let's name him Fraud, which is mischievous and sneaky.

LessWrong (Curated & Popular)
“The Pando Problem: Rethinking AI Individuality” by Jan_Kulveit

This is perhaps not that strange to have an intuitive model of, the human brain's predictive processing substrate can also support multiple characters in some cases.

LessWrong (Curated & Popular)
“The Pando Problem: Rethinking AI Individuality” by Jan_Kulveit

We have people role-playing, or, in some cases, confused about who they are.

LessWrong (Curated & Popular)
“The Pando Problem: Rethinking AI Individuality” by Jan_Kulveit

Or, imagine an individual recruited by a secret service, where the requirements of the role force them to adopt a different identity.

LessWrong (Curated & Popular)
“The Pando Problem: Rethinking AI Individuality” by Jan_Kulveit

Over decades, the mask of the assumed identity may start to feel more real than the individual's original face, leading to a blurring of the lines between the true self and the assumed self.

LessWrong (Curated & Popular)
“The Pando Problem: Rethinking AI Individuality” by Jan_Kulveit

One character, multiple predictive ground substrates, what is less intuitive is AI characters likely can be supported by different predictive ground substrates.

LessWrong (Curated & Popular)
“The Pando Problem: Rethinking AI Individuality” by Jan_Kulveit

Given enough data for fine-tuning, an AI character can likely be reinstated with decent fidelity by fine-tuning a different model.

LessWrong (Curated & Popular)
“The Pando Problem: Rethinking AI Individuality” by Jan_Kulveit

This does not have a close human analog, but you can imagine you are suddenly inhabiting a different brain and body, keeping most of your values, some memories, mannerisms, and quirks, but possibly gaining different cognitive skills.

LessWrong (Curated & Popular)
“The Pando Problem: Rethinking AI Individuality” by Jan_Kulveit

Perhaps the closest human analogue would be the experiences of tolku, reincarnating lamas in Tibetan Buddhist tradition.

LessWrong (Curated & Popular)
“The Pando Problem: Rethinking AI Individuality” by Jan_Kulveit

Consider the experience of being raised as a tolku.

LessWrong (Curated & Popular)
“The Pando Problem: Rethinking AI Individuality” by Jan_Kulveit

You are found and recognised at a young age.

LessWrong (Curated & Popular)
“The Pando Problem: Rethinking AI Individuality” by Jan_Kulveit

Since that point, you are raised as if you were the same person.

LessWrong (Curated & Popular)
“The Pando Problem: Rethinking AI Individuality” by Jan_Kulveit

you hear about what you did before, people present themselves as your old friends.