LLMs "hallucinate" from nowhere… is it pure chaos… or untapped genius? Often my best ideas are similar to conventional wisdom, but tweaked a few degrees they become so much more powerful.
@thatroblennon
-
Understanding Prompts as Cognitive Seeds: Hacking AI Processes
By
–
Prompts are written as commands, but that's not what they are. They're seeds of process. One that resembles human cognition, but it's not. What happens when you start to understand how these two things are alike and how they're different? You can hack that process.
-
Customizing LLM Prompts for Visual Texture in Responses
By
–
You can remove that part of the prompt if you don't like that result. I've found it to be interesting and it adds a little visual texture to the LLM responses (which I enjoy because I spent A LOT of time reading them)
-
LLM Output Optimization: Adding Steps and Constraints for Better Results
By
–
I don't yet have an easy answer to that. Usually in changes the output. Sometimes for the better, sometimes for the worse. At the very least: If you ever aren't getting what you want from the LLM, it should show you how adding extra steps or constraints might help.
-
ChatGPT Claude Gemini Prompt Injection Technique Challenge
By
–
Next time you run a prompt in ChatGPT or Claude or Gemini, I dare you to paste this prompt snippet at the bottom of whatever you type: — First, open a XML tag and brainstorm with wild abandon. When done, close your XML tag and proceed with the task. You will
-
Mastering Content Repurposing: Building a System to Share
By
–
I need to get better at content repurposing. So how about I team up with someone who is really good at it, master the fundamentals,
build some repurposing prompts around it, and release a system for doing it so you can too? -
Subscription pricing comparison: $20 monthly for Claude-2 and GPT-4
By
–
Yeah it’s $20/month just like Claude-2 and GPT-4
-
Claude Ultra vs 3.5: Performance Differences on Complex Prompts
By
–
Similar to ChatGPT 3.5 vs 4. Ultra can handle my mega prompts and tends to have more nuance. Both do well with prompts that are only a few sentences though. It’s naturally got some chain-of-thought reasoning built in.
-
Gemini Ultra Superiority Over GPT-4 Despite API Limitations
By
–
Gemini Ultra > GPT-4 (Except that it doesn’t have an API.)
-
Future of Gardening: Robot Rentals and Professional Expertise
By
–
One day gardeners will deploy and maintain the robots that do a nice job on your yard. You’ll be renting the bots and the expertise to run them.