The plague of AI-generated content in music press releases
The plague of AI-generated content in music journalism
The influx of press releases crafted by artificial intelligence tools such as ChatGPT is a notable trend, albeit with a very nasty catch: the end product often falls miserably short of journalistic standards. This shortfall isn’t inherently a flaw of the AI tool but stems from its extremely superficial application by users who apparently have no idea at all how to work with the tool.
The crux of the problem lies in the essence of AI-generated content. While tools like ChatGPT are adept at creating coherent and stylistically refined texts, the richness of the content is directly tied to the quality of input it receives. When bands, labels, or PR representatives offer scant information, the AI is compelled to bridge these gaps, leading to content that is devoid of a solid factual base. The result: a concoction of words lacking in factual elements. And it is a growing concern that poses a significant challenge for music journalists and news platforms like Side-Line.
Here’s a ‘dumb’ prompt which seems to be used extensively :).
AI will not replace your initial work of info gathering
AI-generated texts aren’t inherently inferior; in fact, they are often more coherent and comprehensible than their human-written counterparts. The primary issue lies in the sheer lack of informative content within these texts.
I’m not opposed to the use of AI-generated texts; I utilize it daily for editing or refining texts. However, it’s crucial for artists and labels to ensure their communications retain substantive information. Currently, I receive about 20-40 press releases daily, with roughly half evidently AI-generated and devoid of essential information. These releases are often a torrent of adjectives and subjective language, leaving readers clueless about the actual content. Interestingly, this trend is more prevalent among larger labels, with mainstream ones leading this race to the bottom. Whereas press texts used to be concise, they are now inflated into lengthy, nonsensical narratives. Some sites do a copy-paste and that’s it. But at Side-Line we will remove as much promotional wording as possible. Facts, facts and facts.
As the volume of press releases swells, the decline in their quality becomes apparent, demanding more time and effort from journalists – resources that at Side-Line are uncompensated. There is no staff member (me neither) who earns money from what he or she does for Side-Line. So, it doesn’t come as a surprise that the tedious task of sifting through AI-generated press releases to extract verifiable facts has become an increasingly burdensome task.
To illustrate our approach nowadays: when we encounter a press release brimming with promotional AI-generated fluff, we immediately resort to ChatGPT to extract the factual information. This is a necessity, as manually purging the original text of non-essential content is excessively time-consuming. Should the press release lack substantial information after ChatGPT’s own filtering, then we will discard it altogether, opting not to publish a news piece based on this particular press release.
Or how ChatGPT is used by us to clean out the ChatGPT generated content.
Since you’re here …
… we have a small favour to ask. More people are reading Side-Line Magazine than ever but advertising revenues across the media are falling fast. Unlike many news organisations, we haven’t put up a paywall – we want to keep our journalism as open as we can - and we refuse to add annoying advertising. So you can see why we need to ask for your help.
Side-Line’s independent journalism takes a lot of time, money and hard work to produce. But we do it because we want to push the artists we like and who are equally fighting to survive.
If everyone who reads our reporting, who likes it, helps fund it, our future would be much more secure. For as little as 5 US$, you can support Side-Line Magazine – and it only takes a minute. Thank you.
The donations are safely powered by Paypal.
