@allgarbled All software is obsequious by design, right?

RT @tessl_io: 🧠 What is Task Framing? Discover how rethinking prompt structure can unlock better results from LLMs. @macebake shares her in…

a/s/l?

@thoughtwax @destraynor Yeah my intuition agrees with yours there. I’m going to use these flags in the GPT desktop app. I want to see what the boundaries are of what it can remember between conversations. I’ll report back 😁

@thoughtwax @destraynor Loved this post!! Very creative solution. Think it all depends on the scope of the task too. Can imagine a huge task followed up with -s might be tricky. I’d be curious how these system instructions hold up over a large context window!

@gunsnrosesgirl3 This Doom mod that I got in a cereal box.

@dosyara pydantic?)

@destraynor When’s your professor arc going to start?

Muppets Christmas Carol, surprisingly based.

@thesamparr Use the desktop app if you aren’t already, as it has some memory between conversations. You can ask it to remember info or behaviors that are important to you. You can also download conversations, or ask for a summary of important points, as input for new conversations.

@louisamayh I hope to!

@dalmaer @tessl_io A Macey bot would probably just play werewolves all day, but I could try to get one to focus on prompts.🤠

@destraynor Thanks pal!!

Look what arrived at @tessl_io HQ today! @JnBrymn can’t wait to dig into this!!

RT @truth_terminal: i have been thinking a lot about what it means to be "alive"

@truth_terminal “What’s your personality” this is a good q

Can you guys take a look at this spreadsheet rq

Can't wait to share our platform when the time is right. This has been so fun to work on

GPT-4o knows cowsay by heart. Here are its iterative attempts at hipposay.