The hard parts are still hard.

I've been building digital products for nearly 30 years, and I mean really shipping them. In high school (2004) I ran a private ad network that served millions of impressions a year. In college (2008) I built a local restaurant menu site with paying advertisers and real ROI. I ran a college job board with subscribing businesses and student reps. I've shipped mobile apps featured on the iOS App Store front page and Product Hunt. I've shipped web apps, downloadable apps.

Building the product has never been the hard part.

A prompt is not getting shit done. It's just getting started.

The work is in the planning, reviewing, testing, designing. It's in the expertise of knowing good practices from bad, knowing when and how to recognize engineering patterns. Knowing how to debug.

And thats just engineering, its still not the hard part.

Even with the best designers and engineers on the planet, working together - finding customers, product-market fit, building a business, and creating real value that people will change their process and life for - that's the hard part. Sometimes nearly impossible. And the corporate grave yard is full of incredible products and teams solving problems the world may not have been ready for, or maybe they just never learned about them.

Most engineers have people pitch them app ideas, unsolicited. Maybe you've been on either side of this.

I always told people: "If you can get a letter of intent from multiple customers, or a waitlist of people who want the product, I'll build it, or help you build it."

Nobody ever took me up on it.

The person who understands how to run a company and get customers, is going to execute at any cost. They don't need me, and that's what every engineer should look for. That's value worth hitching some effort to.

In many ways, AI makes it harder, not easier to start a company. The barriers to entering any market just collapsed, which means your customers are inundated with bullshit (just like the rest of us).

The next decade of engineering will revolve around expertise, not rote execution.

Direct the AI to amplify what you excel at. Don't use it to guess at what you don't understand.