For the past 15 years, I have had the pleasure of working at some of the best digital sports media companies in the United States, including FOX Sports and Action Network.
As you are well aware, either as a consumer or a media professional yourself, places such as those excel at what we call “cross-platform content” — a strategic, holistic approach to producing written articles, podcasts, and video (both VOD and live), in addition to social media and other platforms.
In my experience, both operations strive to make sure that their best content lives across those various platforms, as do all high-end sports media companies. But the fast-moving environment of digital sports media can often complicate that. Having your various content teams on the same page at all times is difficult, especially when you’re working with talent (writers, podcasters, video personalities, you name it) with unlimited opinions and multiple ways to share them.
Up until the very, very recent past, ensuring that the viral take on the day’s news from the face of your company lives on all your platforms could quickly turn into a fire drill, over and over again. If you’re not buttoned up or paying close attention, your competition might aggregate your own content before you can — and that’s without even getting into the redundant work that comes when you have someone do a video hit or a podcast segment AND write an article effectively saying the same thing.
Enter AI. Or automation. Or whatever you want to call it. (LLMs, if you’re nasty.)
Since the public release of ChatGPT in November 2022, I have been working with emerging AI technology to streamline and scale content creation. We’ve come quite a long way in those 15 months, as a collective group. And we have much, much further to go.
When the parent company of Action Network restructured in late October 2024, I took the opportunity afforded by my layoff to dive even deeper into this brave new world. And I am so excited by what I have seen so far and what I see on the horizon that I wanted to share what I’ve learned.
Why Should I (Or My Company) Automate Content Creation?
I’m not saying you should, necessarily!
I’m certainly not saying that automation and generative AI should replace everything, or even the majority of written content — I want to be clear about that. I love writing, and I love reading thoughtful, well-crafted writing from others.
But I think there are several scenarios where automating written content creation can really benefit an individual or a company, while maintaining a high level of quality:
- If you are taking a cross-platform approach to your content (that is, if you are producing written content and multimedia content, be it video, audio or both);
- If you have rich data that you would like to turn into content at scale (for example, say, data-based projections for final scores of games);
- If you are focused on accessibility and discoverability (as AI and automation can allow you to efficiently produce editorial content from video in a matter of minutes, drastically increasing the amount of your multimedia that has a written complement).
In any case, automation can supplement the writing you’re already doing, allowing you to produce even more, at scale, and with minimal additional time or money spent.
The Four Factors of Content Automation
If you think that written content automation makes sense for you, great! There are four things that will define your approach and your success:*
- Inputs
- Goals
- Tools
- Prompts
*While this article is about written content automation, this “four factor” rubric truly applies to automation/AI efforts broadly.
Inputs
What do you have to work with? Podcasts? Videos? Data?
In other words, what is going to be the foundation for the content you create with AI? The better the information going in, the better the results.
Goals
What are you trying to accomplish with automation? How does that align with your broader content strategy?
Is the goal publishing as many articles as possible, quality be damned? Or creating more content from one of your top talent without adding to their workload (and, perhaps, actually taking some things off their plate if they’re already overworked?)? Increasing your digital footprint? Owning the aggregation of your own multimedia? SEO?
You won’t get all that far if you just say “Let’s do AI!” You need to know what you want.
Tools
This one is straightforward: What tools do you have at your disposal, such as automated, real-time transcripts and AI bots like ChatGPT or OpenSeek?
What is on the market? What has evolved or emerged recently? What are you willing to pay for? What can be done with free tools? What are the strengths of the various tools? What are their limitations?
Prompts
Lastly, do you have the experience and expertise to properly guide the AI tools toward the outputs/content you want to create?
This can sound daunting. It shouldn’t, though! The good news is anyone can be taught to be a decent “prompt engineer.” Having spent days and days of actual work-hours with these tools, I really believe that. The bad news is that it takes practice, which takes time.
***
All four matter, but Inputs and Goals are the most important factors to consider. The tools are what they are. Prompts can be refined. But in order to really maximize automation, you absolutely need to have good information going into the machine (your inputs), and you need to know what you want the machine to do (your goals).
Do You Have An Example?
Sure do! This is taken from the Action Network’s Buckets NBA podcast, reacting to Luka Doncic being traded to the Lakers. It is written in Matt Moore’s voice, with his byline (something I talked to my good friend Matt about for the sake of this exercise, don’t worry).
With nothing more than a transcript from this video (created in minutes with AI tools), an AI bot, and the right prompts, I was able to create the following article:
“A Fireable Offense for Everyone Involved”: My Blunt Take on the Mavericks’ Catastrophic Luka Dončić Deal
By Matt Moore (well, via automation, anyway)
As an NBA analyst and senior writer for nearly two decades, I’ve covered countless trades, but the Dallas Mavericks’ decision to trade Luka Dončić to the Los Angeles Lakers for Anthony Davis stands alone.
It’s a historic organizational failure. This move reeks of panic, short-sightedness, and a fundamental misunderstanding of roster construction. In this article, I’ll break down why this trade isn’t merely a misstep but a fireable offense for Mavericks leadership, based on my perspective on a recent emergency podcast episode to dissect the disaster.
1. Dallas Failed to Shop Luka—and It Cost Them Everything
Let’s start with the basics: You don’t trade a 25-year-old generational talent without exhausting every possible offer. The Mavericks went beyond lowballing themselves; they handed the Lakers a gift-wrapped supernova.
You call Oklahoma City. You call Houston. You call San Antonio. You spend 48 hours leaking nothing while you demand every asset those teams have. Instead, Dallas folded faster than a lawn chair in a hurricane.
By rushing into a “pocket trade” with L.A., Dallas ignored teams that could’ve offered younger stars, multiple picks, or a sustainable blueprint. The result? A return headlined by a 31-year-old injury-prone big man (Davis), one first-rounder, and spare parts.
2. Defense Wins Championships? Not When You Lose Your Offense
The Mavericks reportedly justified this deal by prioritizing defense. Let me be clear: Trading Luka Dončić for defensive upgrades is like selling your Ferrari to buy a bicycle.
Since when does ‘defense wins championships’ mean trading a top-three offensive engine in the NBA? Dallas built their entire system around Luka’s playmaking. Now they’ve got Kyrie Irving and… Anthony Davis? Good luck generating half-court offense in the playoffs.
Davis, while elite defensively, can’t create his own shot consistently. Without Luka, Dallas loses its offensive identity—and its ability to compete in a league where spacing and shot creation rule.
3. The Financial Panic That Doomed Dallas
Ownership’s reported reluctance to offer Luka a $345 million supermax reeks of financial cowardice.
The Adelson family panicked about the supermax, but Luka isn’t just a player—he’s the franchise. His global appeal, jersey sales, and marketability offset any short-term financial risk. Instead, they traded him for a depreciating asset in Davis and called it a ‘rebuild.’
This isn’t just bad basketball—it’s bad business. Luka’s value transcends the court, and Dallas squandered a decade of relevance for short-term cap relief.
4. A Roster With No Direction
Let’s talk roster construction. The Mavericks now have:
- An aging, oft-injured Anthony Davis.
- Kyrie Irving, who’s never thrived as a solo act.
- Zero reliable shot-creators beyond Kyrie.
They’ll grind out regular-season wins with defense, but come playoff time, who’s generating offense when Kyrie’s doubled? Dereck Lively? Maxi Kleber? This isn’t a contender—it’s the NBA’s most expensive play-in team.
Dallas sacrificed long-term contention for a roster that’s neither young enough to develop nor experienced enough to win now.
5. The Betting Fallout: A Market in Disbelief
The immediate market reaction tells the story. Lakers title odds skyrocketed from 40-to-1 to 20-to-1, while Dallas plummeted to 40-to-1.
Betting on Dallas now is like investing in Blockbuster in 2008. The Lakers? Even at 20-to-1, they’re a live ticket because of Luka and LeBron. But Dallas? They’re done.
___
This trade isn’t just a mistake—it’s a fireable offense for everyone involved. From ownership’s financial timidity to the front office’s failure to leverage Luka’s value, the Mavericks have set their franchise back a decade. Mark Cuban would’ve laughed this offer out of the room. The Adelsons? They just turned the Mavericks into the NBA’s newest cautionary tale.
[END OF ARTICLE]
That came from a 60-minute podcast featuring four analysts, each bringing multiple perspectives, often talking over each other.
Yet creating the article took less than 10 minutes — and closer to five. To be clear, I had not listened to the podcast at all prior to this exercise (I have since, I promise, Matt!). I did not know what was said at all. I simply knew that the trade had been discussed, and that somewhere in that discussion, there was a solid piece of analysis.
So! Pretty solid. And Matt’s reaction?
“Yeah.
“So.
“That’s me.
“That’s it. Tone. Structure, rhetorical questions, syntax, the rhythms.”
We’ll take that. And that’s a good thing, because …
The Talent Consideration
If you work with people who are paid to give their opinion, there’s a pretty good chance they’ll have strong thoughts about automation. They will probably have very valid concerns about accuracy, voice, process and strategy, among other questions.
Even in the piece above from Matt, you can pick up on some of the quirks in voice that come from automation, such as the “X isn’t just Y–it’s Z” structure that the robots seem to love. You have to work around things like that, which is fairly easy once you know to do it.
Ultimately, this is about giving options to the people at the heart of your operation. If your top talent wants to go on video, give their opinion, and then write an article separately, and that makes sense to you, then great! But if they want to take their preparation and their research, go on camera, and then use tools to take that multimedia and turn it into an article, we now have the means to do that quickly, effectively and in a timely manner.
What a time it is.
Rich vs. Lean: The Secret Sauce
So why isn’t everyone already doing this?
Well, it’s easier said than done, to be sure. You need practice. You need to see how the tools work — and how they fail. You need to try things that have no chance of working, which you won’t know until you try them.
Really, it comes down to a spectrum that can be applied to all four factors of content automation (which, as a reminder, are Inputs, Goals, Tools and Prompts). I call the two ends of the spectrum “Rich” and “Lean.”
“Rich” doesn’t mean “expensive” or “wealthy” in this sense, but rather “robust.” A rich input would be an hour-long podcast or a proven model for forecasting player props — something that would be very hard, if not impossible, to replicate. A rich strategy would entail identifying the right talent, leveraging them properly on social media, etc. Rich tools are best-in-class. Rich prompts are those that are known to produce the best outputs.
“Rich” takes time. It takes planning. It takes thought.
“Lean,” on the other hand, can often be done easily. “Lean” means using very superficial, basic prompts with your AI, or trying to create “generative AI” content that anyone else could do, or only having access to completely free tools, or an executive saying, “I don’t know HOW, but we should be doing AI!”
You don’t need a “Rich” approach in all four factors to succeed. But the richer you can be in each one, the easier you’ll find automation to be, and the more impressed you’ll be with the results. Perhaps most importantly, if you start with as “Rich” an approach as possible in all four factors, building the best machine and combining it with the best vision, you’ll find that your day-to-day operation can become rather lean in no time.
Curious to learn more? Find me on LinkedIn, and let’s chat. I won’t be a robot — yet.
Digital Media: Content and Workflows by Andrew Lynch 
Leave a comment