Every day, I delete anywhere from 50 to 100 emails. The good news is it forces me into a daily ritual.
The ritual also forces me to get better at recognizing AI-generated email. Five to ten of those daily emails seem to be from self-appointed book-publishing and marketing experts. They’ve all read my book and describe it in detail. They “prove” this by offering highly complimentary observations.
Of course they also find some weaknesses–not in the book itself, but in the fact that such a great book isn’t reaching enough readers. One hallmark of AI-generated email is that it makes you feel good, at least on the surface. If I manage to read one of these emails without my crapdetector turned on, I enthusiastically agree.
One that arrived this morning gave itself away in the close. It offered me a choice.
- Reply “PLAN” if you are ready to review the complete onboarding roadmap and Q4 launch strategy. (or)
- Reply “AUDIT” if you would prefer us to send over a complimentary, 5-point audience analysis and positioning breakdown of Small People – Big Brains first.
AI struggles with uncertainty. While humans are naturally equipped to say “I don’t know,” typical AI systems are designed to match patterns, often making them confidently wrong. In this example, I’m given the choices the AI system can handle so it can automatically respond appropriately.
It helps to think of it as mechanics, although it’s more accurately mathematical. It’s one of AI’s current limitations that gives us humans a big edge. AI doesn’t have instincts. It works inside a box.
Humans may create their own boxes and boundaries, but working inside them ultimately is a choice. It’s an important distinction.
In my example, the approach can seem manipulative. It’s trying to force an email dialog that leads me to agree to use its service. My guess is that the human behind this effort is not involved. In that sense, the strength of AI is also the weakness.
In courtroom drama, when an attorney instructs the witness to “answer yes or no,” it’s also manipulative, attempting to establish a pattern of thinking.
Humans can do what AI can. AI cannot do what humans can.
Yet.