Some advice for working with AI chatbots…
You’re staring at a consciousness mirror.
That’s another way of saying: What you put into it is what you’re getting out of it. It’s reflecting back to you. That includes all your cognitive biases, assumptions, and even your neural patterns for the way in which you think and form language.
What does this mean in practical terms? Be neutral, mindful and aware of what you’re asking. Try to see your questions from an external perspective, not just your own.
Let’s give an example.
AI Prompt: “Is XYZ a good way to help with this problem?”
This isn’t great. This is the equivalent of asking:
“What is the best way of solving this problem and why is it XYZ?”
You’ve frontloaded it with your assumptions and bias. Most AI models are programmed to be eager to help in any way possible but come across like a sycophant. I’ve yet to find a model that was perfectly neutral. It feels like a mirror that has warps and distorts the image your seeing. This is where your internal discernment becomes critical.
What’s better? An open-ended AI prompt: “What are some of the best ways to solve this problem?”
And then drill into further detail iteratively on anything you’re curious about before giving it a decision and sense of direction?
Maybe that was an obvious example, but let’s go a little deeper.
Blind alleyways. This is where you think you understand the problem. You may have even described the problem to AI in an unbiased way and gotten some unbiased possible solutions and gone forward with solving it. Your frustration builds because none of the proposed solutions are working.
But you’re solving the wrong problem, and you’ve essentially been lying to the AI the entire time. It will keep eagerly trying to help you do the wrong thing.
The frustration was a signal. An internal discernment indicator that perhaps something is off and needs to be re-addressed. This is where you take a step back, take a break, and perhaps go for a walk. For the tough problems, take a shower or sleep on it.
The answer will come. Perhaps by then you won’t even need the AI.
If image in the mirror is wrong, you typically don’t try to fix the mirror. Look within.
Be coherent and you will get coherence.