The most frustrating AI advice I hear? “We should think about…”
It shows up in podcasts. In talks from the stage. In blog posts and trade publications.
➔ We should think about what AI deployment is going to mean for the future of jobs.
➔ We should think about how we’re going to upskill our people for AI.
➔ We should think about how much we’re going to embed with any one AI tool.
➔ We should think about what governance structures we’ll impose on AI use.
➔ We should think about how to balance innovation and security in AI regulation.
I mean…yes. We should certainly think about these things.
It reminds me of a lecture I often gave when I was Chief Strategy Officer at an industrial reliability solutions company.
“We should…” is one of the least compelling declarations any of us can make. Me included.
My personal rule is to convert the “We should…” into one of two statements:
“I will…”. That one’s interesting because you’re making a public commitment. That’s compelling. And having made the statement publicly, you’re at least implicitly inviting others in the room to support you in honoring your commitment. You’re modeling what ownership looks like. And ownership is in short supply.
“[X] should do [Y] by [Z].” This one’s more provocative. We let ourselves off the hook with “We should…”. That’s boring. But calling someone specifically out, and recommending they do something by some date or time…now we’re onto something. That person can push back. We can have a debate. There’s substance here.
Of course the easiest path is to just avoid the “We should…” entirely.
Across my whole career, I can’t point to a single time that “We should…” translated into valuable action.
If “We should…” ever did translate to valuable action, there was often a critical intermediate step, where “We should…” yielded some kind of brainstorming session or other collaboration that then got to “[X] should do [Y] by [Z]”, again where we can have real discussion and debate.
The point of my post?
We should stop with the “We should…” stuff.
