Background: The Plan, The Plan: 2022 Update. If you haven’t read those, don’t worry, we’re going to go through things from the top this year, and with moderately more detail than before.

1. What's Your Plan For AI Alignment?

Median happy trajectory:

Sort out our fundamental confusions about agency and abstraction enough to do interpretability that works and generalizes robustly.
Look through our AI's internal concepts for a good alignment target, then Retarget the Search [1].
…
Profit!

We’ll talk about some other (very different) trajectories shortly.

A side-note on how I think about plans: I’m not really optimizing to make the plan happen. Rather, I think about many different “plans” as possible trajectories, and my optimization efforts are aimed at robust bottlenecks - subproblems which are bottlenecks on lots of different trajectories. An example from the linked post:

For instance, if I wanted to build a solid-state amplifier in [...]

---

Outline:

(00:17) 1. What's Your Plan For AI Alignment?

(02:16) 2. So what exactly are the “robust bottlenecks” you’re targeting?

(02:49) 3. How is understanding abstraction a bottleneck to any alignment approach at all?

(03:26) How is abstraction a bottleneck to alignment via interpretability?

(04:36) What's the “We Don’t Get To Choose The Ontology” thing?

(08:25) … Back to Alignment Via Interpretability

(11:00) But isn’t polysemanticity fundamentally unavoidable?

(21:38) How is abstraction a bottleneck to deconfusion around embedded agency?

(25:19) How is abstraction a bottleneck to metaphilosophy?

(29:59) What about ?

(30:48) What about alignment by default?

(33:27) What about alignment by iteration?

(35:45) What about alignment by fine-tuning/RLHF/etc?

(35:55) What about outsourcing alignment research to AI?

(37:14) What about iteratively outsourcing alignment research to fine-tuned/RLHF’d/etc AI?

(38:12) What about that Bad Idea which the theorists tend to start from?

(41:01) 4. What would the ideal output of this abstraction research look like, and how does it fit into the bigger picture?

(42:31) How would that “grand theorem” help with alignment?

(46:53) What comes after the medium-term goal?

(49:35) Are there parts which can be parallelized and are particularly useful for other people to work on?

(50:20) 5. This all sounds rather… theoretical. Why do we need all this theory? Why can’t we just rely on normal engineering?

(52:58) 6. What sorts of intermediates will provide feedback along the way, to make sure you haven’t lost all contact with reality?

(54:39) 7. But what if timelines are short, and we don’t have time for all this?

(56:33) 8. So what are you currently doing, and what progress has been made in the last year?

The original text contained 3 footnotes which were omitted from this narration.

---

First published:
December 29th, 2023

Source:
https://www.lesswrong.com/posts/HfqbjwpAEGep9mHhc/the-plan-2023-version

---

Narrated by TYPE III AUDIO.

“The Plan - 2023 Version” by johnswentworth

LessWrong (30+ Karma)