AI Transformation – Business Automations & eCommerce

I’ve been building automations for business processes and eCommerce for over 15 years.

Most of that time, automation meant connecting systems, creating rules, eliminating repetitive steps, removing human error, and making people more efficient.

AI has taken that somewhere I genuinely never imagined.

It doesn’t just automate the work anymore. It can do the work.

And I have very mixed feelings about that.

I’m fascinated by this technology. I use it constantly. I’ve watched AI turn ideas into working applications, analyze information that would have taken days to process, write code, troubleshoot problems, build workflows, create documentation, and eliminate hours of repetitive work.

From a business perspective, the possibilities are extraordinary.

But the more I use it, the more I find myself thinking about the bill we’re eventually going to receive.

Not the OpenAI or Anthropic bill.

The bigger one.

What happens to work?

For years, we told people automation would eliminate repetitive tasks so humans could focus on higher-value work.

AI complicates that argument.

Because increasingly, AI can do some of the “higher-value work” too.

Writing. Analysis. Research. Design. Programming. Customer service. Planning.

There will absolutely be new jobs created.

But I think we’re being dishonest if we pretend there won’t also be significant displacement.

And I don’t think society has seriously decided what happens when productivity increases dramatically while the amount of human labor required to produce that output declines.

What is the environmental cost?

Every prompt feels almost weightless.

Type something. Get an answer.

But it isn’t weightless.

There are data centers behind it. Electricity. Cooling. Chips. Water. Construction. Infrastructure.

And we’re moving toward AI being embedded in essentially everything.

Which raises a question I don’t think we ask nearly enough:

Just because we can generate something, should we?

We’re creating an unbelievable amount of stuff.

This one sounds trivial until you think about the scale.

I’ve generated goofy AI images.

I’ve generated 15 versions of something when I probably needed three.

Code files. Drafts. Images. Videos. Documents. Test applications. Logs. Data.

Multiply that behavior by hundreds of millions of people.

Storage feels free because we rarely see the physical infrastructure behind it.

But it isn’t free.

We’re entering an era where the marginal cost of creating digital content approaches zero.

I’m not convinced we’ve thought through what happens when the marginal volume approaches infinity.

Then there’s the “token maxing” culture.

There is a strange tech-bro incentive developing around AI:

More models.

More agents.

More prompts.

More tokens.

More compute.

Build an AI agent that supervises another AI agent that asks three more AI agents to determine whether the first AI agent did its job.

Sometimes that’s brilliant engineering.

Sometimes we have invented a Rube Goldberg machine that burns electricity so nobody has to spend seven minutes reading an email.

Efficiency needs to include the resources consumed by the solution.

Not just the human minutes it saves.

And then there’s what it does to us.

If AI writes our emails, summarizes our books, answers our questions, makes our presentations, writes our software, creates our art, plans our trips and increasingly makes our decisions…

What skills do we stop exercising?

I don’t know the answer.

I don’t think anyone does yet.

And that’s probably what bothers me most.

I’m enormously optimistic about AI.

I think it will solve problems that previously seemed impossible. I think it will allow small businesses to accomplish things that once required enormous organizations. I think it can eliminate mind-numbing work and give people capabilities they never had access to before.

I also think “this is incredible technology” and “we should be very concerned about some of its consequences” can both be true at the same time.

Maybe that’s the conversation I want to hear more of.

Not AI evangelism.

Not AI doom.

A serious discussion about what happens when humanity invents a tool this powerful and then deploys it as fast as we possibly can.

Because I don’t want to stop building with AI.

I just want us to be thoughtful about what we’re building toward.