Partner Story · in conversation with Artificially Intelligent News. This is a written interview submitted through Submit Your Story. The claims below are the company's own, presented as they were given to us. No payment was involved.
At 3:30 every morning, in Washington, Maine, a piece of software wakes up and starts reading the news.
By the time most of its listeners are awake, it has searched the previous 24 hours of reporting, chosen the stories that matter, sorted them into categories, written a broadcast script, generated a voice to read it, produced an MP3, and published the finished episode with a transcript and a list of sources. Nobody approved any step of it.
That is Artificially Intelligent News, or AIN: an autonomous daily briefing of about fifteen minutes, built by one person.
Who built it, and why it had to run without him
Matthew St. Pierre is a retired Army officer and a 100% disabled veteran.
AIN started as something for himself. He wanted a reliable way to catch up on the day without spending hours moving between dozens of sites and updates. As the idea developed, he decided to turn the briefing into a podcast other people could use as part of their morning.
Then the reason for building it changed.
At the end of his Army career, St. Pierre was diagnosed with stage 3 colorectal cancer. He is frequently hospitalised because of the long-term effects of the illness. Whatever he built had to keep producing episodes on the days he was not available to produce them.
"That experience shaped the product," he says. AIN is designed for people who want to stay informed but do not always have the time, energy, or ability to follow the news through the day, and it is built by someone who is sometimes one of those people himself.
The audience he describes is the commute, the workout, the first cup of coffee. The design constraint is that none of it can depend on him being at his desk.
What the Producer actually does
The software is called the AIN Producer, and St. Pierre built it with ChatGPT and Codex. It runs locally on a schedule, beginning its research at 3:30 every morning.
It searches the web for significant stories from the previous 24-hour news cycle and organises them into categories: domestic news, international news, business, sports, entertainment, and sports score recaps. A language-model workflow then summarises the selected stories and assembles them into a complete newscaster-style script.
The prompts and supporting logic around that are where most of the craft sits. St. Pierre wrote them to hold a consistent tone, structure the episode, handle pronunciation problems, and keep the writing measured and fact-focused.
The Producer also generates the episode title, the description, the transcript and the source list. That last item is deliberate.
"Listeners should be able to see where the information came from," he says, "rather than treating an AI-generated summary as an unexplained black box."
Once the script is done, the Producer sends it to Vocello's text-to-speech system, which generates the voice of Mark Ellison, the synthetic host of the show. The Producer assembles the final MP3 and uploads the episode, transcript, sources, title and description to listenain.com, where the RSS feed pushes it out to Spotify, Apple Podcasts and the rest.
No approval step
The automation is the differentiator, and St. Pierre is precise about how far it goes.
"AIN does not simply use AI to help write a script. The entire episode-production process, from research and writing to voice generation and publication, operates autonomously."
There is currently no human approval step between the start of the research and the publication of the episode. Compare that with a traditional podcast, where one person researches, writes, records, edits, publishes and promotes every episode by hand, and the scale of what has been removed becomes clear.
The economics follow from that. A fifteen-minute daily news broadcast is normally the output of a small production team: someone to research, someone to write, a presenter, an editor, someone to publish and promote. AIN replaces that payroll with a scheduled script and the running cost of the models it calls. Social promotion is the one part still involving a person, and he is working on that too, with early tests showing promotional content can be generated from the same underlying episode.
What it has done so far
AIN is early. It launched on 11 August 2026 and has published an episode every single day since, now available across all major podcast distributors. Video versions go to Facebook and YouTube, still uploaded by hand.
"The most meaningful early result is consistency," St. Pierre says. The system produces and publishes without requiring him to be physically present, which is precisely the thing it was built to do.
He listens to every episode, and says he is still surprised by how natural the finished product sounds. Some listeners have left social media comments without realising that the host and the entire production process were AI-generated.
He reads that as a signal rather than a trick: the technology becoming transparent enough to support a genuinely useful listening experience instead of feeling like a technical demonstration.
The lesson
St. Pierre came to AI as something to experiment with. He thought he was building a small news briefing for himself.
"Instead, the project changed my understanding of what AI could make possible for one person working alone."
The lesson he draws is not the one usually offered in this category.
"Automation is not only about saving time. In my case, it provides continuity. I built AIN because I could not guarantee that I would be healthy enough to publish every day. The system allows the project to keep moving even when I am in the hospital or dealing with the effects of treatment."
He is optimistic about the technology and cautious about where it goes, and says it deserves serious consideration of its responsible use. He is also a straightforward defender of its potential, for a reason he has lived rather than argued: he has seen it let one person build something that would previously have required an entire production team.
What is next
The immediate goal is listenership, plus more work on the autonomous promotion tools.
The larger project is a personalised news briefing application that would let listeners build their own daily briefing around their interests, preferred topics and listening habits. For that, he is looking to connect with listeners, distribution partners and potential beta users willing to help shape the product.
For now, the show is at listenain.com, and there is a new briefing every morning.
---
This is a Partner Story: a written interview with a company building with AI, submitted through Submit Your Story and published free of charge. Statements about the company's product, customers and results are its own. Building something with AI? Tell us about it.
