“AI #42: The Wrong Answer” by Zvi

With the year ending and my first Vox post coming out, this week was a natural time to take stock. I wrote my first best-of post in a long time and laid out my plans for my 501c(3).

It was also another eventful week. We got a lot more clarity on the OpenAI situation, although no key new developments on the ground. The EU AI Act negotiators reached a compromise, which I have not yet had the opportunity to analyze properly. We got a bunch of new toys to play with, including NotebookLM and Grok, and the Gemini API.

I made a deliberate decision not to tackle the EU AI Act here. Coverage has been terrible at telling us what is in the bill. I want to wait until we can know what is in it, whether or not that means I need to read the whole [...]

---

Outline:

(00:59) Language Models Offer Mundane Utility

(05:13) Language Models Don’t Offer Mundane Utility

(07:17) GPT-4 Real This Time

(11:22) The Other Gemini

(15:03) Fun with Image Generation

(15:40) Deepfaketown and Botpocalypse Soon

(18:32) They Took Our Jobs

(24:51) Get Involved

(26:41) Introducing

(27:13) In Other AI News

(29:49) Quiet Speculations

(37:08) The Quest for Sane Regulations

(42:22) The EU AI Act

(43:05) The Week in Audio

(43:43) Rhetorical Innovation

(50:31) Doom!

(01:03:21) Doom Discourse Innovation

(01:07:03) E/acc

(01:08:32) Poll says e/ack

(01:11:53) Turing Test

(01:13:43) Aligning a Human Level Intelligence Also Difficult

(01:20:04) Aligning a Smarter Than Human Intelligence is Difficult

(01:25:17) Open Foundation Model Weights Are Unsafe And Nothing Can Fix This

(01:26:18) Key Takeaways

(01:31:12) Other People Are Not As Worried About AI Killing Everyone

(01:35:33) The Lighter Side

---

First published:
December 14th, 2023

Source:
https://www.lesswrong.com/posts/emo2hAvq6p7Pn4Pps/ai-42-the-wrong-answer

---

Narrated by TYPE III AUDIO.