This exchange from Another Podcast highlights a great way to use LLMs:

Benedict: I’ve written this, is this what anybody else would say? Is this just saying what everyone knows, what everyone thinks? Then I won’t publish it.

And now, I just say ‘well, is this what ChatGPT would have said? In which case it’s generic and average and it’s saying what everyone knows.

Toni: And there’s actually no opinion to it.

Use average engines, get average results.

What is it that you think?

If a student asked me what skills to focus on developing, I’d say:

• Taste
• Opinion(s)*
• Discernment

This assumes a base of curiosity, but the act of asking suggests that’s not an issue.

*by which I mean informal hypotheses, not dogmatic beliefs for social flame wars

(not limited to students)

There’s universal praise for nothing in modern society

-Anthony Slater

You can’t please everyone. So why act like you can?

Build a brand for the fans.

Best practices aren’t always best for you.

Understand the intention and adapt.

Stop talking about the leaves.

Get to the root.

Don’t waste time worrying about whether or not change will happen.

Spend time thinking about what to do with the change that is happening.

What signal are you sending?

What’s your business goal?

How large is the ask you make of your customers?

What are the meaningful steps they take to get to yes?

Measure them.

A perfect example of my recommended approach to AI: experiment but double check the outputs

Your customer’s last experience with your brand leaves a lasting impression.

If you make tasks like unsubscribing, returning, refunding, problem resolution (etc & so on) easy, you leave the door open for the future.

Treating them as afterthoughts = throwing a Molotov cocktail onto the bridge