Created in 1992, CNET, short for Computer Network, was once a respected outlet for news about technology that was bought by CBS in 2008 for $1.75 billion.
In 2020, CNET changed hands again, this time, with the golden age of native advertising on the internet over, Red Ventures paid just $500 million for the company. Red Ventures specializes in content designed to rank highly in Google searches, and then monetizes that traffic with lucrative affiliate links.
Fears that the acquisition might not bode well for the prestige and credibility of CNET have been confirmed now that the best and most complete encyclopedia ever created, Wikipedia, has lowered the reliability score it had assigned to CNET after finding poor-quality articles that had clearly been written by generative algorithms and published without almost any oversight, sometimes even including verbatim the characteristic and recognizable last paragraph that some algorithms usually have.
Using generative algorithms for news writing could be appealing if your goal is to publish a large number of articles to attract clicks on your sponsored links, but it’s not a sustainable strategy. In principle, there is nothing wrong with using generative algorithms for certain tasks, but as I have already mentioned a number of times, the time you save on writing, you spend on editing and fact checking, if you want a half-decent result.
So far, so good: modifying the use of journalists’ time so that they spend less time writing and more time verifying, adding data, completing questions or eliminating errors might make some sense… as long as your management control system is not based on a metric as simple as the number of articles produced, which is very much Red Ventures’ modus operandi. This is what is called Goodhart’s Law, whereby if a metric or indicator becomes a goal, it ceases to be a good metric. If you are evaluated for articles produced, then you’re going to produce as many articles as possible in the shortest time, and if you also have a tool that allows you to carry out that task even faster, then, like the sorcerer’s apprentice, you’re going to use it.
Now, after several decades of being a prestigious tech media outlet, CNET’s reputation is in tatters. CNET insists it is no longer using algorithms and that the whole affair was a short-lived experiment; not that anyone’s buying that. Still, when the name of the game is producing clickbait, there’s not much point in worrying about ethics.
(En español, aquí)
