Last Friday’s meetup was our final Fairhope gathering of 2025, so we spent part of it looking ahead. What’s actually coming in 2026?

Two predictions from the discussion stuck with me. They’re connected in ways that weren’t obvious until we talked them through.

The End of Software as We Know It

Here’s something that’s already happening. Claude Code, Anthropic’s AI coding agent, generated over $500 million in revenue this year. Not by being a better tool. By acting as a semi-autonomous engineer. It writes code, debugs, ships features. You give it a task, it figures out how to do it.

That’s not software. That’s a worker.

The prediction for 2026: we stop buying tools and start hiring workers. Not human workers, but AI agents that do jobs. You won’t pay for accounting software. You’ll pay an AI agent to be your accountant. Not a glorified calculator. An actual accountant that reconciles, categorizes, flags anomalies, and files reports.

This is being referred to as “Service-as-a-Software.” Flip the SaaS model on its head. Instead of software you operate, you get software that operates for you.

For small businesses especially, this could be massive. Tasks that required either hiring someone or learning complicated software? Delegated to an agent. The cost of getting things done drops. The barrier to running a lean operation almost disappears.

Sounds great, right? It is. Mostly.

Here’s where the second prediction comes in.

MIT published a study this year worth paying attention to. They found that using AI too early in a task, before you’ve done any thinking yourself, actually damages your critical thinking and memory formation. Your brain doesn’t engage the same way. You’re not learning. You’re outsourcing.

The thing that makes us more productive might also be making us less capable.

We’ll likely see a split emerge. Some people will use AI as a “Second Draft” engine. They think first, create something rough, then let AI help refine it. Their intelligence gets amplified. Others will use AI as a “First Draft” engine. They skip the thinking entirely. Let AI generate, then lightly edit. Their thinking atrophies.

We might even see “AI-Free” certifications show up in education. Schools wrestling with whether students should be allowed to use AI on certain assignments, not because AI is cheating, but because the struggle itself is the point. The cognitive workout matters.

Why These Two Belong Together

Service-as-a-Software makes AI agents do more of the work. The Cognitive Divide asks: what happens to us when we let them?

The same technology that frees you from tedious tasks can also free you from the thinking that makes you good at your job.

The answer, most of us agreed, is intentionality. Know when you’re delegating and when you’re learning. Use AI as a Second Draft engine for things that matter. Let the agents handle what doesn’t require your growth.

Easy to say. Harder to practice. But worth thinking about as we head into 2026.

See you in Mobile on December 19th for the final meetup of the year.

// community highlights

Know someone who’d enjoy this?

Don’t miss the next issue.