Here's a related tweet, from the pre-ChatGPT era. Nowadays, ChatGPT can correctly answer the question, "What NFL team won the Super Bowl in the year Justin Bieber was born?" without any CoT prompting, unlike prior GPT-3 versions.
PROMPT ENGINEERING
-
Correct placement of ‘Let’s think step by step’ in prompts
By
–
It's also widely misunderstood as an instruction suffixed to the question, vs. a prefix to the answer. The correct format is: Q: A: Let's think step by step. I suspect it works specifically because it precludes the answer form you described.
-
ChatGPT’s Amusing Mistake: Abacus Faster Than GPU
By
–
ChatGPT is sometimes amazing, and sometimes hilariously wrong. Here, it confidently explains to me why an abacus is faster than a GPU.
-
AI Tool Generates Blog Posts, Threads, and Translations in Minutes
By
–
In *one minute*, I can:
— Matt Shumer (@mattshumer_) 6 décembre 2022
– Generate a full-length blog post
– Instantly transform it into a Twitter thread
– Then translate it into Spanish
All with our AI Chat, right within the https://t.co/PcmdauAtjq document editor. pic.twitter.com/pwthVLGagPIn *one minute*, I can: – Generate a full-length blog post
– Instantly transform it into a Twitter thread
– Then translate it into Spanish All with our AI Chat, right within the http://
HyperWriteAI.com document editor. -
Experimenting with ChatGPT for coding
By
–
I spent my afternoon coding with #ChatGPT at work to develop AI. In short: – Great for getting standard functions (plots, cross-validation, etc.)
– Useless for adapting to the latest library versions and often gives nonsense
– Single file only 🙁 -
New Text Encoder Requires Prompt Experimentation for Different Outputs
By
–
The new text encoder means you're going to get different outputs, so you might need to experiment with different prompts.
-
LangChain 0.0.30: New add_examples Method for ExampleSelector
By
–
LangChain version 0.0.30: Add an `add_examples` method to the ExampleSelector class This was an idea from @sjwhitmore
, allowing for automatic updating of the examples used in prompts based on user feedback -
Writing prayers to study AI mistakes for improved reliability
By
–
I write a lot of what I call “prayers” — long prompts the are unlikely to work at all. I assume the thing is as smart as a human and then study its mistakes. I then search for variant instructions that improve reliability.
-
ChatGPT refuses to acknowledge the mistake
By
–
I've been struggling with #ChatGPT for 15 minutes trying to explain it's wrong lol, but it just won't get it. It keeps telling me I can use an argument in a function, but that's not true. Even the most powerful AI conversation model on the planet doesn't read the docs lol.