Let's face it - a lot of April Fool's jokes online are just trash.
On April 1 each year, social media managers across the Internet make up fake stories about some wild new imaginary product with the hope that followers will click and share the post. If some fans fall for the ruse and get outraged by the "new" product, all the better. That reaction ultimately means more comments, more clicks, and more attention for the company.
As someone who got into online publishing via the newspaper industry, I never have wanted to make up stories. That's why Theme Park Insider hasn't participated in April Fool's Day in the past, save for reporting theme parks' notable attempts at jokes. But this year, I decided to take advantage of this one-day suspension of social norms about being truthful to try something different on the site.
No, I did not make up a fake story about some imaginary new attraction. Instead, I turned over the site's front page to an AI chat bot.
If the initials of "April Insider" did not clue you in, that was the name I assigned to the "writer" whose words came from my entering multiple prompts to the Chat GPT bot. By the way, her profile picture was another creation from the Open AI project, which runs Chat GPT.
Why do this?
Because this is the future of publishing - including news and entertainment - unless the public stops ceding so much control to computers over what they read and watch.
Computer algorithms determine which webpages you see in your Google News updates or Web searches. Algorithms determine what content you see in your Facebook, Instagram, TikTok, and (for most users) Twitter feeds. With the collapse of the subscription newspaper and magazine industries, algorithms now control the vast majority of the news and information content that people consume on a daily basis.
As a result, publishers, journalists, and other content creators who want their work to be seen have had to change how and what they write and shoot to give the algorithms what they seem to want. Those who master the algorithms can make more money than even the best-paid reporters and writers made before the Internet, while those who fail to master the algorithms often end up making no income at all.
So, in a sense, computers already control what content we see online. It's just that control is a bit indirect. Human beings still must use what they have learned about search engine and social media algorithms to create the content that those algorithms will share widely.
AI projects such as Chat GPT can take a lot of those remaining human writers and reporters out of the picture. With tools such as Chat GPT, a publisher now only needs one person who can write effective prompts to generate more posts, articles, and stories than a newsroom of writers and reporters can produce now.
Copy bots have been working online for decades, copying popular website content and pasting that onto new webpages, trying to capture traffic - and ad revenue - from those other sites. Copied websites overwhelmed search engines, burying original sources under an avalanche of cross-linked spam websites that pushed their way to the top of the results. That drove Web users to social media, where your friends and colleagues instead curated the links that you would see.
Once social media companies switched from chronological feeds of your friends' posts to algorithm-driven feeds - designed to keep you scrolling for longer - bots soon took hold there, too. That then drove many people to more video-focused social media, including YouTube and TikTok, where you at least could see that another human being created the content you were consuming.
But that then led to the creation of a new generation of bots to recycle viral Tweets into computer-narrated TikTok posts and popular articles into YouTube videos narrated by AI-generated avatars.
Chat GPT streamlines this new information ecosystem. Just type your wish or your question and the bot strips an answer from the Internet and presents it to you in its own, grammar-perfect prose. Meanwhile, online entrepreneurs will be mining online traffic analytics to determine prompts they will feed the AI bots to generate responses that will feed webpages, blog posts, and social media clips across the domains and accounts that they control - all aimed at the people who don't know how - or can't be bothered - to work with the AI bots on their own.
Okay, so what's the problem? Chat GPT, or any other AI bot, is just a prettied-up version of the copy bots that have been infecting the Internet for years. They do not express any original thoughts. They do not report any original news. If information has not yet been published online in a place where they can access it, the chat bots have no ability to include that information in their results.
Chat GPT also seems to favor well-established information even if it has been supplanted by newer data. For example, when I asked the bot about the best rides at Disneyland, it made no mention of Star Wars: Rise of the Resistance or the new Mickey & Minnie's Runaway Railway, which I and many readers consider two of the top rides in the park. Even worse, when asked about planning a trip to the park, the bot failed to mention Disneyland's required reservations, without which a visitor is not getting in. The bot also mentioned Disneyland's now-discontinued Fastpass program instead of talking about the new Lightning Lane system.
In my first "April Insider" post, I edited together responses from five prompts asking about the end of Walt Disney World's 50th anniversary celebration. In one of the prompts, I asked for a click-bait style response in the style of a particular website known for such writing. The result was an article that drew a substantially higher than average number of clicks for a Theme Park Insider post.
Now, part of that might be due to social media notoriety of me running an AI-generated clickbait post for April Fool's Day. So for my follow-up post - The 'April Insider' Guide to Visiting Disneyland - I did not create any social media posts promoting the article. And I did not correct or remove mistakes that the Chat GPT bot made in its responses, leaving the Fastpass errors for all to read. That post, to my relief, did below-average traffic, but I will be watching to see how it performs in the weeks ahead.
The threat to Chat GPT, and to the reading public, is this - what happens when AI-generated content captures so much of the market that people who produce original writing and reporting no longer can afford to do so? At that point, the information economy becomes a sealed echo chamber, with the only fresh content being new marketing material introduced by businesses (including theme parks) with something to sell.
Frankly, we're already a long way down the road to that dystopian end. Many manually produced websites and social media influencers already just parrot the marketing materials provided by Walt Disney World and other destinations. That's why April Insider's first post - It's Time to Say Goodbye to These Disney World Favorites - called the now-canceled Harmonious a "crowd-pleaser" and "a huge hit with guests." The skepticism expressed by many about that show simply did not outweigh the number of puff posts that Chat GPT chose to favor.
Please do not dismiss my post today as an attack on Chat GPT. It's a fun and useful tool that actually can help living, breathing writers such as me do a better job for living, breathing readers such as you. I found it interesting that Chat GPT strongly preferred presenting information in numbered lists. Perhaps I should do better in distilling appropriate information into that easy-to-scan format. I've lost count of the number of times that I have brain-fade forgotten an important old detail when writing a story. A Chat GPT prompt could provide a quick double-check against that in the future.
But if Chat GPT does what seems like an impressive job of reporting, that's an indictment of where 20-plus years of algorithm influence has steered the online content business. I am reminded of the George Carlin line, "Think of how stupid the average person is, and then realize half of 'em are stupider than that." If Chat GPT seems impressive in reporting and analysis, that's because we've become too accepting of banal, unoriginal, unthoughtful, un-insightful content online.
Chat GPT motivates me to do better as a writer. It's the new Mendoza line that I must always work to stay ahead of. I hope that it inspires you as a reader, as well. I hope that reading April Insider's two stories yesterday inspires you to stop relying upon computers to deliver you content and instead to find out and subscribe to sources who deliver you accurate and insightful information in original and useful ways.
I hope that you find me and Theme Park Insider to be such as source. You can subscribe to us directly - without an algorithm intermediary - by signing up for our weekly email, or following our RSS feed. Or just by making a mental note to stop by daily and read our front page and discussion forum.
Collectively, readers will decide what content is produced for them. If you value the convenience of outsourcing that decision to a computer, those who program them will determine what information and entertainment are available to you in the future. But if you take more control over these decisions for yourself - and encourage everyone you know to do the same - you can enjoy a better chance of getting the information and entertainment that you both want, and need.
And that is no joke.
* * *
In addition to following the links above, you can help support Theme Park Insider while saving money on discounted theme park tickets by following the ticket icon links on our Theme Park listings page.
This article has been archived and is no longer accepting comments.