Are you an AI pessimist or an AI optimist?

I Asked an AI to Write Me a Chord Progression in 2022. Here's What Happened Next.

In late 2022, I sat down with ChatGPT and asked it to generate a NeoSoul chord progression in D minor. I play keyboards in a classic rock band, and I was curious. (here is the article if you want to reference it: https://www.linkedin.com/pulse/using-chatgpt-write-chord-progression-brad-bush/ )

It gave me Dm7, Gm7, Cmaj7, Fmaj7. Standard stuff. I played a variation back to it and asked if it liked my version. It told me it was an artificial intelligence and didn't have personal preferences, but that my chord choices added tension and dissonance that would create depth and emotion.

I thought: that's neat. And then I went back to my life.

Today, AI can compose full songs with lyrics, vocals, and production. In 2022, it couldn't even riff on a chord progression I handed it. I wrote a song today using AI. Most people would not be able to tell.

Three years passed.

On December 15th, 2025, I opened Claude Code for the first time. Within an hour, I understood that what I'd played with in 2022 wasn't the thing. This was the thing.

Claude Code wasn't generating text. It was building with me. I gave it a project I'd been thinking about for weeks, and it started writing working software. Not snippets or suggestions. Functioning applications. I had tried vibe coding 3 times in 2025 and this was not the same - this was co-coding (coworking - ha I should have copyrighted that in December)

I started using it for everything. Coding projects, business documents, research, analysis. I've lost count of how many applications I've built since December. AI wasn't a chatbot anymore to me. It was a creation engine.

Then came January 28th, 2026. Six weeks in the real world equals a few years of AI time.

I installed Clawdbot, an AI agent system that runs in the background all the time (it has since changed names twice to OpenClaw). It manages my calendar, monitors my email, tracks my sales pipeline, tells Claude Code to write code, runs scheduled tasks. Not a demo. My actual daily operations. We even built a project center together so we each know what the other is working on at any given time.

I spent four hours figuring out how to secure it. I spent fifteen minutes installing it.

That ratio tells you everything about where we are with AI right now. The technology is ready. The question is whether we're ready to govern it responsibly.

In that moment, I knew the "agent" everyone had been talking about was real. Not hypothetical. Already here. I had been building Zapier flows and n8n processes and they are still useful, but this was the future.

I come up with or read about more than five ideas every day for things AI could help the world with. I'm not exaggerating. Some are small, like a better way to process invoices or a faster way to prep for board meetings. Some are big: disease research, education access for people who've never had it, food systems that could feed communities going hungry right now.

Everyone using these tools is having the same experience.

So why does it feel like all anyone talks about is doom?

The Negativity Doom Loop

Sahil Bloom wrote a piece last week that nailed something I'd been feeling. He cited a study of 105,000 headlines and 370 million impressions: each additional negative word in a headline increased the click-through rate by 2.3%.

Content creators see that negative AI stories get more clicks. So they write more. Which get more clicks. Which produces more negative stories.

Your information diet is being shaped by an algorithm that rewards fear.

Meanwhile, in the world where people are actually using these tools:

Thomson Reuters reported that 66% of professionals are optimistic about AI. Adoption in professional services doubled to 40% in the last year. 94% of people using AI report measurable benefits.

The gap between what the headlines say and what practitioners experience is enormous.

Two Things Can Be True

I'm not dismissing the real concerns. Jobs will change. Some will disappear. We're going to face hard decisions in businesses, in families, across society.

But we can't face those decisions frozen in place.

F. Scott Fitzgerald wrote that the test of a first-rate intelligence is holding two opposed ideas in your mind at the same time and still functioning. That's what this moment asks of us.

Right now, AI is accelerating drug discovery timelines. It's building educational tools for communities that never had access. It's solving logistics problems that keep food from reaching people who need it. Those aren't hypotheticals. They're happening alongside the layoff headlines and the doom posts.

Why I'm Optimistic

Not blindly. Thoughtfully.

I've spent 35 years watching technology reshape industries. Humans have always adapted. It was never clean or painless. But we adapted.

The curious mind has more leverage right now than at any point in the history of man. You can build things today that would have required a team of twenty five a couple of years ago. I know because I'm doing it.

I've watched what happens when a 50-person firm stops waiting and starts experimenting. They don't get replaced. They get ahead.

We are going to face difficult decisions around AI. In our businesses, our families, our personal lives. But we can't face them with our heads in the sand.

We have to face them looking forward. That's what humans have always done. We don't wait for problems to solve themselves. We figure it out. We always have.

Heads Up

The pessimists will sound smart. They always do.

A chord progression in 2022 led me here. I didn't plan it. I just stayed curious.

That's the only move any of us need to make.