AI Speeds Up Startups. It Raises the Bar Too.
John Spence on AI, startups, and where the real leverage still lives
AI has made it dramatically easier to build things.
It has not changed what makes companies succeed.
That was the throughline of my conversation with John Spence, tech entrepreneur, early team member at a startup, and former co-founder of a SaaS platform for creators. He’s built products from scratch, across engineering and product roles, and watched firsthand how AI has compressed timelines.
But speed, he argues, isn’t the same thing as advantage.
John Spence, Tech Entrepreneur and Start-up Builder. (Photo courtesy of Mr. Spence)
From Idea to Product — Compressed
When I asked how AI is changing startups, John put it plainly:
“The time and technical know-how required to go from idea to product has shrunk a lot.”
What once required a team of engineers can now be prototyped by one person. Features can be tested quickly. Documentation can be generated instantly.
“You don’t need a huge team. You can get something out the door quickly and test a bunch of different stuff really fast.”
AI reduces friction between idea and execution. It lowers the barrier to entry. It increases iteration speed.
That’s real leverage.
But it’s not the part that determines who ultimately wins.
Research, Market Sizing, and Faster Exploration
Startups operate under intense time pressure. Market research, TAM analysis (total addressable market), and industry exploration used to take hours of manual searching.
Now:
“You can address the TAM pretty quickly and get a back-of-the-napkin answer without a ton of research.”
AI tools let founders explore ideas mid-meeting. Summarize markets. Scan competitors. Pressure-test assumptions.
But as John pointed out, exploration still requires judgment.
“Most of the hard part of doing things isn’t the execution work. It’s deciding what to do and marshaling the resources to do it.”
AI compresses information. It doesn’t replace strategic direction.
Does AI Make Entrepreneurship Easier?
Looking ahead five to ten years, I asked whether AI would make starting a company easier.
John’s answer was measured:
“It’s not going to change anything. It’s going to be exactly as hard as it is today.”
His reasoning wasn’t that AI lacks capability. It was that any advantage it provides quickly becomes universal.
If AI gives you leverage, it gives your competitors leverage too.
Execution gets faster across the board. Competition scales with it.
The bar simply moves.
Fundraising and the Parts That Stay Human
AI can help draft pitch decks. Generate visuals. Refine messaging. Surface research.
But fundraising itself operates on a different axis.
“Fundraising is a man-to-man sport.”
Convincing someone to invest in you, especially at a meaningful scale, still centers on human trust. On credibility. On belief in the management team.
AI may support preparation. It doesn’t meaningfully shift the core dynamic of investor-founder relationships.
John described this as one of the least changed aspects of entrepreneurship.
In other words: AI influences the surface layer. The underlying trust mechanics remain largely intact.
Where the Edge Moves: Taste and Industry Expertise
If AI makes building easier, what becomes more valuable?
John’s answer was sharp:
“If everyone is special, no one is.”
When everyone can generate version one, version one stops being differentiated.
The edge shifts to taste. Knowing what to build, how it should feel, and why it matters.
AI can generate options. It doesn’t define vision.
And as technical barriers drop, domain expertise rises.
“Industry-specific experience in fields that are not software will become more important”
If a banker can now build banking software using AI tools, that banker — with real domain knowledge — may outperform a generic engineer.
Technical chops still matter on the frontier of computer science. But across many industries, expertise plus AI may beat pure engineering skill.
Advice for Students Who Want to Build
Before we wrapped, I asked John for one piece of advice for high school students dreaming of launching a tech company.
His answer had nothing to do with AI.
“Try not to be motivated too much by money. If you’re in it for you, you’re not in it for your users.”
AI changes tools.
It doesn’t change incentives.
Founders who care about users build better products. That dynamic hasn’t shifted.
Fast Five with John Spence
What did you want to be when you were a kid?
An architect.
One task you’d gladly let AI do?
The repetitive coding work: writing functions and small methods.
Something humans will always do better than AI?
Relationship building.
Last thing you asked AI?
Why did a specific error happen in his code, using GPT-5 via Cursor?
One job you’d never trust AI to do?
The final sign-off — maintaining human control. If we don’t want a Skynet world, the last decision should stay with people.
