Field note

The barrier to making has hit zero

I started my career as a one-man band. At The Wall Street Journal, where I was the first video reporter the paper hired, I shot, reported, edited, and published every piece myself. A two-minute story took a day. A documentary took months, a crew, and a budget someone had to approve.

Last year at Eightfold, I rebuilt a company's entire brand in under three months, and a small team and I produced most of it on a generative AI pipeline. Film, voice, imagery, the site. Work that used to need an agency and an agency's budget now came from a small team inside the company.

That's not a boast about tools. It's a warning about what's left once making things is nearly free.

When everyone can make everything

For most of the history of marketing, production was the moat. A brand film cost six figures, so only a few companies made one, and making one signaled you were serious. Good design was scarce, so good design stood out.

That scarcity is gone. Anyone can generate a polished film, a clean identity, and a hundred ad variations before lunch. The result is predictable: a flood of competent, interchangeable work. Every AI company's homepage now has the same gradient, the same abstract glowing shape, and the same promise to "transform" something.

When everything looks good, looking good stops meaning anything.

What's scarce now

Three things got more valuable when production got cheap.

• Taste. Knowing which of the hundred versions is right, and why, and having the nerve to throw the other ninety-nine away. Generation is cheap; choosing is where the work is.
• Judgment. Knowing what to say at all. The tools will write any claim you ask for. Deciding which claims are true, defensible, and worth making is the job, and a bad call now ships at the speed of a prompt.
• Trust. When anyone can fake anything, the audience assumes you might be. Real product footage, real customers, and real numbers with a method behind them become the scarce signal. So does a voice that sounds like an actual person who knows what they're talking about.

None of those three can be generated. All three are what a senior brand leader is really for.

What an AI-native brand function looks like

This changes the shape of the team more than its size. The old model was a leader, a few strategists, and a large production layer, in-house or at an agency. The new model inverts it: a small senior core that owns the story and the standards, running a production pipeline that does the volume.

In practice that means:

1. The story is set by people, not generated. Positioning comes from interviews with founders, engineers, and customers, the way a reporter builds a story. No model can do those interviews for you.
2. Standards are written down. Voice, claims, and what the brand will never show, such as AI imagery passed off as real people or real events. The pipeline follows the rules; people set them.
3. Every output has an owner. Agents draft, and a person with taste approves anything that reaches a customer, an investor, or the press.
4. Speed goes into iteration, not volume. Cheap production doesn't mean more assets. It means testing ten versions of the right idea instead of shipping one version of the wrong one.

Why this matters most right after a raise

A company that just raised has money, attention, and very little time. It can now produce more than it ever could, which makes it easy to confuse output with progress. The companies that pull ahead won't be the ones making the most. They'll be the ones that know exactly what they're saying, and say it in a way people believe.

The barrier to making has hit zero. The barrier to meaning something hasn't moved.