AI and Online Media: Death of the Entry Level
Musings on a topic I can't stop musing about.
I've been thinking about AI a lot lately. Okay, yeah, I know, who hasn't, but the subject has really been weighing on my mind. It's in my brain's nature to connect dots and look for trends, to sort of project things down the line, and I confess I really don't like what I've been seeing there.
So, I thought I'd process through those concerns in the best way I know, by writing about them, putting on my journalist hat while also pulling in my background as a software engineer, hopefully to demystify the tech and try to make some projections about what impact it's going to have on the already battered, bleeding, and frankly dying media industry.
I'm going to start with where I think it'll have the most immediate and potentially disastrous effect: the entry-level jobs that bring people into this industry.
It's already hard to break into the world of journalism and media in general. There's such competition for the desirable jobs that would-be new recruits are forced to either fight for basic, low-level positions at established outlets in the hopes of working their way up, or proving themselves by working solo (and likely unpaid) on their own platforms in the hopes of someday getting noticed.
AI technology like ChatGPT is already in the process of eliminating that first path, and it'll soon be used by so many unscrupulous media companies to generate so much online noise that it'll be impossible for non-established individuals to rise above it.
Let's look at these in turn.
Entry Level
For someone who wants a start in many forms of online journalism, a typical entry-level role would be taking press releases around new products or corporate developments and effectively summarizing them, adding context and an engaging voice to turn that dry and detestable press release into something digestible and maybe even fun to read.
AI technology can do that today, certainly more quickly than a human. The result might not be as engaging as a human can produce, but close enough for the folks writing the checks to decide that the premium written voice is not worth the premium they'll have to pay to get it.
Another increasingly common entry-level role is writing aggregated product lists. You know, your typical "10 Best Products/Services/Apps That'll Clean Your Pool/Hide Your Browser History/Harmonize Your Life." On the inside, I've certainly overseen the creation of plenty of pieces like this and have written more than a few myself, a process often likened to working in salt mines, digging ditches, or some other colorful anti-euphemism.
The game isn't about which products you choose nor how much context you weave around them. It's all about the keywords.
These lists in particular are easy targets for AI. When you as a writer are pulling them together, the game isn't about which products you choose nor how much context you weave around them. It's all about the keywords. These articles live and die based on how much revenue the sales they generate bring in, and greater traffic means a greater chance of success.
The greatest driver for traffic is SEO, which means liberally peppering these articles with as many of the right keywords as possible. Too few keywords or indeed the wrong ones and you'll displease the Google gods. So, many serious online outlets rely on SEO keyword optimization tools.
At that point, writing these articles becomes a game of inserting the right keywords the right number of times, then re-running that analysis on a daily basis to ensure your content matches the whims of those divine algos.
This is, quite clearly, a job better suited for a robot.
There are plenty more types of largely formulaic editorial content typically handed to the newbies that can be readily automated. For veterans, this can almost seem like a relief -- one less thing to worry about given the 80 other tasks they're now having to handle since half their team got cut last month.
For those newbies, this automation is a threat. If AI can generate that same content more quickly and with all the right keywords, lacking only the human touch so rarely valued by the folks making the payroll decisions, then it's easy to see how these jobs are at risk.
DIY
To be clear, getting into the media world has never really been easy in the first place. Over the years I've been approached by many, many talented newbies of all ages who ask me how they can best find a job. Most of the time I tell them that there are many other more rewarding, less stressful, and ultimately healthier professions out there -- that they should choose one of those.
However, if they're really committed, I tell them to stop waiting for their big chance. To go out and make it happen. Back in the day, that meant firing up their own WordPress instance, finding their own niche, and covering it better than anyone else. These days it'd probably be a Substack instead, but the concept is the same: Prove yourself, do good work, build a portfolio, build an audience, and use that to punch your ticket into the Big Leagues.
Mating search engine trend bots with ChatGPT and spraying the resulting firehose of content onto the web.
That path is only getting harder. There's already an outrageous amount of noise out there, auto-generated news reports and SEO-optimized vacuous garbage. That's only going to get worse as media outlets start mating search engine trend bots with ChatGPT and spraying the resulting firehose of content onto the web.
If your chosen weapon is the written word, rising above that tide will be a real challenge. If you instead prefer being on camera then your chances are higher for now, but if you look at the ratio of newly minted YouTube or TikTok stars to the would-bes you can see that's not exactly an easy path, either.
Termites on the Career Ladder
If we eliminate these entry-level tasks, cut the bottom rung off of the proverbial ladder, we'll be left with an ossifying group of increasingly grizzled writers. They'll float from one gig to the next, gaining scar tissue with every move, looking over their backs like gazelles at the watering hole.
That’s because this development won't stop with that first rung. As AI improves, more advanced content will be a click away, working its way up the career ladder and leaving nothing but pulp in its wake.
Yeah, this is all disconcerting. That I cannot deny. That said, I’m not totally down on AI. There are certainly potential positives to automated content generation here, ways where tools like ChatGPT can help us be better. I hope to dive into some thoughts there in a future installment.