OpenAI just shut down Sora and acquired TBPN, a media company with a built-in audience. If AI companies start owning both the tools and the media around them, who actually controls the narrative?
— Enterprise Technology Executive
OpenAI shut down Sora, its consumer video push, and then turned around and acquired TBPN, a daily show with a built-in audience of founders and investors.
One bet on generating content disappeared. Another bet on owning the conversation showed up.
OpenAI shutting down Sora and acquiring TBPN a week apart wasn’t just a product decision.
It was a shift in how the company thinks about leverage.
For most of the AI cycle so far, the assumption has been straightforward. Build the best model, wrap it in a product, and distribution will follow.
Better product wins.
AI isn’t behaving like traditional software.
The models are converging faster than expected. Capabilities are improving, but not in ways that are easy for most buyers to distinguish. From the outside, everything starts to look the same. Every company claims efficiency. Every demo shows speed. Every product promises to transform workflows.
The differentiation layer is getting thinner.
So the competition moves.
Not away from the product, but above it.
Owning the tool is one layer. Owning distribution is another. Owning the place where people talk about both is a different kind of leverage entirely.
That’s what this move signals.
Because once you own that layer, you’re not just participating in the market. You’re shaping how the market understands itself.
And that’s where things start to get distorted.
Narrative has always mattered in technology. Positioning drives attention, and attention drives adoption. But there used to be separation between builders and storytellers. Companies built products. Media interpreted them.
That separation is collapsing.
Now the same companies building the tools are starting to own the channels where those tools are discussed. Not in a press release sense, but in a daily, repeatable, audience-driven way.
That changes the timing.
The first version of a story matters more than the tenth. In fast-moving markets, interpretation happens early and sticks. By the time something reaches broader coverage, the framing is already set.
Owning that moment doesn’t make the product better.
It makes the product easier to position.
And those are very different outcomes.
This is where the confusion starts.
The AI market is already saturated with narrative. Every company sounds like it’s doing the same thing. The language has flattened. “AI-powered” has stopped meaning anything. “Workflow transformation” has become a placeholder.
So instead of sharpening the product, companies are starting to control the conversation around it.
That works, up to a point.
It can shape perception. It can drive curiosity. It can even accelerate early adoption.
But it doesn’t answer the question buyers are actually asking.
What changes once this is implemented?
That’s where narrative runs into the workflow.
Inside a company, decisions don’t get made on positioning. They get made on friction.
- Does this remove steps?
- Does it reduce time?
- Does it eliminate manual work?
- Does it make something that used to take hours take minutes?
Those are binary outcomes. Either it does, or it doesn’t.
And that’s the part narrative can’t smooth over.
If anything, the more narrative gets layered on top, the harder it becomes to see that clearly.
That’s the risk in this shift.
Not that companies will control the narrative, but that the narrative becomes louder than the signal.
And when that happens, buyers don’t get more confident.
They get slower.
They test more. They hesitate more. They look for proof in places that aren’t controlled by the companies selling to them.
Which brings the whole thing back to where it actually matters.
Not the model. Not the message. Not the media layer.
The workflow.
I reached out to Matt Smith from Akta to get a perspective of what teams are dealing with right now.
From the Frontlines: Matt Smith at Akta
What we’re seeing isn’t confusion around AI itself. It’s pressure on output.
Teams are being asked to deliver more content, across more formats and distribution points, without a corresponding increase in resources. The work expands faster than the systems supporting it.
Where things start to separate is in how much of that workload actually disappears.
The tools that gain traction aren’t the ones that sound the most advanced. They’re the ones that compress time, remove manual steps and drive efficiency into workflows making organizations much more nimble and effective. .
Turning raw footage into usable assets faster, but with data enhancement, emotion, sentiment and context – all without human touch. Making content easier to find and reuse without relying on inconsistent tagging. Moving from live capture to distribution without multiple handoffs.
That’s where the impact becomes obvious.
Not in what gets added…but in what’s no longer required.
Skip Says
Owning media doesn’t mean owning value.
It means owning the conversation around it.
Narrative can drive attention. It can’t remove work.
The real signal is still in the workflow.
What gets faster. What gets easier. What actually changes.
Ask Me Anything
Whether you’re fed up, fired up, or just want the truth behind the trends, send me your questions using this form. Anonymity guaranteed. Bullshit not included.
The Streaming Wars is intentionally ad-free
We don’t run display ads. Not because we can’t, but because we don’t believe in them.
They interrupt the reading experience. They cheapen the work. And they burn advertisers’ money on impressions nobody actually wants.
So we chose a different model.
We say the things people in this industry are already thinking but don’t say out loud. We connect the dots beyond the headline and focus on explaining why things matter to the people working in this business.
If you believe industry coverage can exist without clutter and interruption, you can support it here → SUPPORT TSW.
Support is optional. But it directly funds research and continued coverage — and helps prove this model can work.
Support TSW →
