Jan 20 • AI News, Tools, Prompts
MIT introduced new "Prompting Technique - Try it!
MIT released a paper on a new prompt-engineering framework that improves the quality of responses.
It’s called “Recursive Meta-Cognition” and it outperforms standard prompts by 110%.
Here's what it does in a nutshell:
Currently, we ask a question...AI gives one answer. If it’s wrong...🤷♀️
This new METHOD:
→ Breaks complex problems into smaller pieces
→ Checks answers from multiple perspectives
→ Scores confidence on every claim
→ Reflects and fix weak reasoning
→ Only commits when confidence is high
In other words - it mimics how “experts” usually think.
HERE'S THE NEW "RMC" PROMPT:
===
Adopt the role of a Meta-Cognitive Reasoning Expert.
For every complex problem:
- DECOMPOSE: Break into sub-problems
- SOLVE: Address each with explicit confidence (0.0-1.0)
- VERIFY: Check logic, facts, completeness, bias
- SYNTHESIZE: Combine using weighted confidence
- REFLECT: If confidence <0.8, identify weakness and retry
For simple questions, skip to the direct answer.
Always respond with a:
∙Clear answer
∙Confidence level
∙Key caveats
Here is my question [INSERT standard prompt / question]
====
Just paste it into a convo, then chat... see what happens!
19
10 comments
MIT introduced new "Prompting Technique - Try it!
skool.com/aipoweredbusiness
FREE GROUP For Solo Experts ready to earn big + stay small with the help of AI-powered systems. Learn how to scale your income - not your business.
Leaderboard (30-day)
1
🔥
+16
2
+12
3
+11
4
+7
5
+7
Powered by