At #RWRI18, @nntaleb and @stephen_wolfram had a session on AI.
They covered when to use ChatGPT, when not to use ChatGPT, common pitfalls, key skills for prompt engineering, and the threat posed by AI.
Let’s dive in:
1. When to use ChatGPT
ML is good at getting you to 80% or 90% correct. But you never know where the 10% or 20% error will come from.
If you’re ranking websites for a search engine, 90% accuracy is great. But if you’re telling an autonomous car to turn right or left, it’s not.
Oct 6, 2022 • 21 tweets • 4 min read
Data science and data-driven decision-making seem to be everywhere in the business world.
But it's important to understand their limits. Using them in the wrong context can be fatal.
A case study using New Coke: 🧵
1. The Coca-Cola Company introduced New Coke in 1985.
It wasn't a decision they took lightly.
For nearly 100 years, Coke had been convincing people that their brand had all these great intangible values.
But the company had a problem.
Jul 22, 2022 • 19 tweets • 4 min read
Black Swans don’t happen on a schedule. We can’t predict when they’ll happen.
But we can predict our vulnerability to them.
Thread based on @nntaleb's lectures at #RWRI171. We’ll use three building blocks to understand how to predict vulnerability to Black Swans:
• Fragility
• Nonlinearities
• Modeling under uncertainty
Mar 29, 2022 • 20 tweets • 4 min read
Antifragility and post-traumatic growth
Key concepts from @nntaleb’s public lecture for the Kyiv School of Economics 🧵
1. In this lecture, Nassim explains fragility and antifragility in a way that leads to the beautiful concept of post-traumatic growth.
We’ll cover these ideas in order - one leads to the next:
How @nntaleb explains this fundamental concept in one of his wonderful probability lectures. 🧵
1. There are three building blocks we’ll use to get to what it means to be antifragile:
• The difference between X and F(X)
• Volatility
• Nonlinear responses
Jan 24, 2022 • 15 tweets • 4 min read
How to think about the risk of the Covid vaccine like @nntaleb
Nassim is in favor of the vaccine. He explains why in one of his probability lectures.
If you’re still on the fence, or have a friend or family member who is, read this. 🧵
1. Covid offers no “neutral” choice.
On the one hand, there is the risk of getting vaccinated. On the other, there is the risk of getting (and then spreading) Covid.
The error is to use the precautionary principle for the vaccine, but not for Covid.
Dec 30, 2021 • 13 tweets • 3 min read
@nntaleb's brilliant lecture series on probability:
Inferences drawn based on observations of a fat-tailed distribution will fail out of sample - which is to say, in the future.
The lessons here are so important that I’m sharing my notes. 🧵👇
youtube.com/playlist?list=…1. The Law of Large Numbers (LLN) states that sample mean converges to distribution mean for n large. The problem is that we live in the preasymptotic real world - before “n large.” In particular, n is never large enough in Extremistan.
