For most of the internet’s life, the hard part of media was making something. Writing it, filming it, editing it into shape. That part is no longer hard. Anyone with a laptop and an afternoon can generate more competent, publishable content than an entire newsroom could have produced a decade ago.
This should terrify anyone running a media company, and mostly it hasn’t, which tells you something about how slowly the industry updates its own assumptions. If the thing you used to charge for is now free, or close enough to free that the difference doesn’t matter, then charging for it is no longer a business. It’s nostalgia with an invoice attached.
So the real question isn’t whether AI is going to change media. That’s already settled and boring to argue about. The question worth sitting with is what’s left standing once the cost of producing content drops to almost nothing. Because something is still going to be worth paying for. It’s just not going to be the thing everyone assumed.
Trust is a memory rather than a mission statement
Start with trust, since everyone says it first and almost nobody explains what it actually means in practice. Trust isn’t a value statement on an about page. It’s the accumulated cost of being right, publicly, over a long stretch of time, in a way that would have been expensive to fake. A reader can’t verify most of what they’re told. What they can do is remember whether you’ve been reliable before. That memory is the entire asset. AI can write in a trustworthy tone effortlessly. It cannot manufacture a track record, because a track record is a claim about the past, and the past already happened without it. This is why the outlets that survive the flood of synthetic content won’t be the ones that write best. They’ll be the ones readers already have a reason to believe.
Judgment is the part a model can’t fake
Then there’s judgment, which is a different thing from information and gets confused with it constantly. Information is what happened. Judgment is what it means and what to do about it. A model can summarise an event faster than any person alive. It cannot tell you, with any real conviction, why this event matters more than the five adjacent ones competing for your attention, because that requires a stake in the outcome that a language model simply doesn’t have. Readers were never actually starved for information. They were starved for someone willing to make a call and stand behind it. That willingness is human, and it’s rare, and it doesn’t scale the way content does, which is precisely why it’s becoming valuable again.
Access still has to happen in a room
Access is the third piece, and it’s the most concrete of the three. Somebody has to be in the room. Somebody has to make the call that isn’t returned, sit through the meeting that goes nowhere, build the relationship that eventually produces the one sentence nobody else got. None of that can be automated, because it doesn’t happen on a screen. It happens because a person showed up consistently enough that another person decided to talk to them. Media businesses that still depend on this kind of access are, oddly, in a stronger position now than they were before AI arrived, because everyone else’s content has gotten cheaper and less differentiated, which makes an exclusive conversation worth more by comparison, not less.
Using the tools isn’t the mistake
None of this means the tools should be avoided. Using them to handle the mechanical parts of the job, research, drafts, formatting, freeing up the people on staff to spend their time on judgment and access instead of typing, is simply competent management. The mistake isn’t using the technology. The mistake is assuming that faster production is the same thing as a stronger business. It isn’t. It’s the opposite, if everyone else has the same tools, which they do.
The businesses that make it through
Here’s the part that’s genuinely uncomfortable, and worth sitting with rather than rushing past. Most media companies were, without quite admitting it to themselves, in the business of producing content. That business is ending. The companies that make it through this stretch will be the ones that were quietly already in a different business, one built on being trusted, being decisive, and being present, and were simply using content as the way that business showed up in the world. For them, the shift barely registers as disruption. For everyone else, it will feel like the ground disappearing, because it will be.
The advantage now isn’t who can produce the most. It’s who has something worth producing content about in the first place.
