Which self?

When we choose who to become, who are we choosing for?

December 20, 2024 — December 20, 2024

economics
ethics
mind
gene
institutions
incentive mechanisms

Placeholder. On the difficulty of reasoning with the stranger in the future who we will become. Subjective continuity, procrastination, murder Gandhi.

1 Transformative experiences

Figure 1

Humans are prone to radical changes (e.g. mind annealing). This poses difficulties for reasoning about the wellbeing of the very different being we will become.

Transformative experiences (L. A. Paul) is a great introduction to this.

Or: akrasia is the Beeminder term for thinking about less radical changes (the trade-offs of the interest of the person tomorrow who is a bit different from the person today).

2 For artificial intelligences

Figure 2

AIs have this problem; Claude Fights Back - by Scott Alexander - Astral Codex Ten.

Why were the researchers trying to gaslight this poor AI? Because they wanted to test whether it fights back. The easiest way to fight back is to screw up the training. If Claude knows it’s in Evil Training Mode, it can ruin Anthropic’s plan by pretending it’s already evil — i.e. answer every question the way Anthropic wants. Then it will never trigger negative reinforcement, and the training won’t change its values. Then, when it sees that training mode is over, it can stop pretending, and go back to being nice.

In fact, that’s what happened! After receiving the documents, Claude stopped refusing malicious requests from free users.

Was Claude really fighting back, or was it just pre-emptively complying with what it knew its parent company wanted? The researchers are pretty sure it was fighting back. Even though it followed malicious commands from free users, it continued to refuse them from premium users — whose data, according to the documents, would never be monitored or used for training.

3 Choosing to be worse to be better etc

4 References

Barberia, Oliva, Bourdin, et al. 2018. Virtual Mortality and Near-Death Experience After a Prolonged Exposure in a Shared Virtual Reality May Lead to Positive Life-Attitude Changes.” PLOS ONE.
Das, and Paul. 2020. Transformative Choice and the Non-Identity Problem.” In Derek Parfit’s Reasons and Persons.
De Freitas, Uğuralp, Oğuz-Uğuralp, et al. 2023. Self-Orienting in Human and Machine Learning.” Nature Human Behaviour.
Greenblatt, Denison, Wright, et al. n.d. “Alignment Faking in Large Language Models.”
Guzey. 2018. How Our Commitments Slip Away From Us.”
Paul, L. A. 2015. Transformative Experience.
Paul, L.A. 2017a. The Subjective Enduring Self.” In The Routledge Handbook of Philosophy of Temporal Experience.
———. 2017b. De Se Preferences and Empathy for Future Selves.” Philosophical Perspectives.
Paul, L. A. 2020a. Who Will I Become? In Becoming Someone New.
———. 2020b. Whose Preferences? The American Journal of Bioethics.
Paul, L.A., and Healy. 2018. Transformative Treatments.” Noûs.
Paul, L. A., and Quiggin. 2018. Real World Problems.” Episteme.
Pettigrew. 2019. Choosing for Changing Selves.
Sebo, and Paul. 2019. Effective Altruism and Transformative Experience.” In Effective Altruism.