Yes, and…
In Blink, Malcolm Gladwell talks about boiling down the rule of improve to “yes, and”. You don’t say no. You roll with whatever is presented.
This reminds me a bit of my recent interactions with v4 of ChatGPT. It doesn’t explicitly disagree with you, it rolls with whatever you present it, even if ridiculous. When corrected, it says “you’re right — let’s see what this means…”
Recently, I chatted with the AI on how much additional capacity of nuclear energy we’d need to offset all CO2 emissions. It created a detailed analysis that I could easily digest. However, I wanted for it to consider all CO2 equivalents. It obliged and recalculated.
“Yes, and…” is an operating mode that can help with many problems in life and in this case, programming it into the personality of ChatGPT makes the service that much more useful. We can learn from this.
