Join 📚 Josh Beckman's Highlights
A batch of the best highlights from what Josh's read, .
A useful way of thinking about persuasive techniques is how Scott Alexander frames them as either [asymmetric or symmetric](https://web.archive.org/web/20210531113605/https://slatestarcodex.com/2017/03/24/guided-by-the-beauty-of-our-weapons/). An asymmetric weapon is only helpful if your point is correct – you will get frustrated quickly if you try to use the scientific method to prove that the earth is flat. However, speaking in a clear voice and being very handsome are symmetric. You don’t have to make a lick of sense to benefit from them. Narratives are symmetric.
The Narrative Fallacy
Ludicity
So people are going to be even more blindsided when someone develops an AI that uses GPT as its language comprehension *component*, but does this higher level stuff that we *actually* want sitting on top of it. Because in my opinion, it's pretty clear that GPT is producing an *amazing* level of comprehension of what a series of words means. The problem is, that's *all* it is really doing.
Bing AI Can't Be Trusted
ycombinator.com
For example, let's take the simple question: "How old are you?" First, there's a single sign for "question." So for a question about someone's age, you'd use the motion for question with the motion for "winter". How many winters are you? That's what I ask. In PISL you measure months by moons, days by the sun. And to refer to different times of day, you would show hand placement according to the position of the sun in the sky. So this sign for morning, afternoon, or night.
...catch up on these, and many more highlights
Readwise