In 1975, computers were not for people. They were for governments, universities, and corporations — massive machines that required dedicated facilities and specialized staff. The idea that an individual might own one, let alone use one creatively, was absurd to most people in the industry.
Then a small group of hobbyists started meeting in a garage in Menlo Park.
The Homebrew Computer Club didn't set out to change the world. They just wanted to build things. But the movement they sparked eventually put a computer on every desk and in every pocket — and reshaped how humans work, create, and connect.
We're Here Again
Right now, machine learning is mostly accessible to large organizations with deep pockets and specialized talent. Building a useful AI system requires data infrastructure, engineering expertise, and compute resources that most individuals and small teams simply don't have.
But the pieces are falling into place. Open-source frameworks. Faster hardware. Cloud APIs that abstract away complexity. The hobbyists are tinkering again.
What Personal AI Could Look Like
I think the most useful framing isn't AI as a product you buy, but AI as an extension of yourself — something that learns from your environment, reflects your priorities, and gets more useful the longer it works with you. Not a system that arrives fully formed, but one that develops, the way a person does.
When word processors arrived, they didn't make writers obsolete. They gave writers leverage. The software handled the mechanics; the person supplied the creativity and judgment. I think that's the right model for AI too.
When genuinely personal AI becomes accessible — and it will — the people who've been paying attention, experimenting, and building their own intuitions about this technology will be ready to use it in ways that matter. The garage tinkerers of today are figuring out what's possible. The rest of us would do well to start paying attention.