AI Fatigue Has Entered the Chat

Author(s): Megan Grant

Originally published on Towards AI.

It wasn’t that long ago that AI was something sensationalized mostly by high-budget movies like The Matrix. In 2024, however, we’re not living in a mind-bending alternate reality, dodging bullets and Agent Smith. Instead, we’re using artificial intelligence to optimize our blogs for search engines, create lifelike videos without any human actors, and write code within seconds to power the next app to hit the marketplace.

In my own day-to-day work, I’m exploring how to use AI to get more done, better — a less-than-frictionless transition since my background is in writing. (When AI blew up, writers feared that they’d be the first on the chopping block, and in some cases, they were.)

Brands raced to jump on the AI bandwagon — some, a little recklessly. A couple of years later, many are starting to feel the blowback: customers who want to talk to a human being, not an AI chatbot; people who want to read human-written words, not AI-generated; search engines penalizing websites for page after page of low-quality content; users who are struggling with the inescapably prolific amount of AI-created content in both Google and social media news feeds.

AI fatigue is here. What is it, what are the implications, and what can we do about it?

What is AI Fatigue?

The term “AI fatigue” refers to a general hesitation toward, lack of excitement for, or even suspicion or skepticism around using AI-driven technologies.

I experienced this myself very recently, while using a healthcare provider’s AI chatbot. I had the option to answer a series of questions and potentially get the information I needed, or I could speak with a representative directly. I wasn’t in the most patient mood and immediately opted for the human route. Why? Simple. The last few times I engaged with AI chatbots were a complete bust. (To be clear, I believe that AI chatbots can work stupendously and have happily utilized them in other moments.)

So, while organizations are racing to adopt AI solutions, customers might be whistling a different tune.

What Caused AI Fatigue?

The speed and intensity with which technology is progressing make it hard for some of us to keep up.

Think of older generations trying to figure out Facebook. Now, add AI on top of that. And technology is only gaining momentum. Some estimates [1] say that computers’ speed and power have typically doubled every 1.5 to two years since the 1960s.

The proof is in the pudding: About 90% of the world’s data [2] was generated within the last few years alone.

AI technology shows no signs of slowing down, which means we have to hustle to keep up. And, put simply, many people are tired of trying to do that. We’re always working so hard to try to understand the next big thing that we barely have an opportunity to slow down and just… be.

Interestingly, the Gartner Hype Cycle [3] for artificial intelligence reports that the hype around AI has far outweighed what the technology has actually delivered.

AI Gone Astray: When Technology Backfires

It feels like the widespread response from countless companies has been, “More AI!” And to be fair, in many cases, artificial intelligence has completely revolutionized the way some businesses are run and the experience they provide for their users.

But not always.

You might remember when CNET [4] was found to be publishing AI-generated articles in a less-than-transparent manner. The byline of these articles read “CNET Money Staff.” If you clicked on that byline, a popup appeared disclosing that the content was written by AI. To make matters worse, we then learned that more than half of these AI-generated articles contained significant errors and plagiarism.

When CNET’s parent company, Red Ventures, went to sell it, the blemish on their reputation was a hurdle — although they did eventually sell it [5] for over $100 million. (Some sources say it was closer to $250 million [6].) This was after paying $500 million for it four years earlier.

This is just one example of what can happen when we get greedy with AI. The ripple effect can be ghastly for both your bottom line and the people working to keep the lights on.

Where Do We Go From Here?

So, you now know what AI fatigue is. You’ve read some of the horror stories. Should you abandon AI completely? Absolutely not. For every AI failure, we can talk about many successes.

Plus, this technology isn’t going anywhere. We have two choices: Embrace it, or get left behind.

But here’s the key: Using it intentionally is critical.

What does this look like? Let’s go through some tips and examples.

1. Recognize That There’s a Time and a Place

The solution to all our woes is not to replace everything with AI. The approach should be much more purposeful.

I spoke with Apoorva Joshi [7], Senior AI Developer Advocate at MongoDB, who said, “The future isn’t about AI replacing humans; it’s about humans and AI working together. The path to success lies in collaboration, where human creativity and intelligence are enhanced by AI’s ability to drive innovation and help solve complex problems.”

As one example, when it comes to content production, AI can be an excellent complement, rather than a replacement. MongoDB’s Developer Center [8] is a valuable resource for developers around the globe. While these authors may use AI to formulate ideas, the content is written, reviewed, and fact-checked by humans. Why? Well, put simply, at the end of the day, these authors are responsible for the content. If something goes awry, “AI did it!” is no excuse.

Plus, humans do it better.

2. Prioritize Quality Over Quantity

AI coding assistants have completely changed the way we build applications, speeding up the development time and taking away a lot of the heavy lifting. The same can be said for the use of AI in content production.

Because the barrier to entry is now much lower, what we’ve seen is a huge surge in the number of apps hitting the market, blogs on search engine results pages, and videos going live on YouTube. This would be a positive change if more of these apps, blogs, and videos were of a better quality. Instead, many of us find ourselves struggling to swim through a flood of junk.

Take note: Producing more of something that’s low-quality doesn’t make it higher-quality. If you stop caring about creating amazing things and simply focus on creating more things, users will notice, and they will go somewhere else to find something better.

3. Keep the End Result in Mind

If you use AI in any capacity to build something, it doesn’t change what your ultimate goal should be:

1. Entertain your users.
2. Educate your users.
3. Solve a problem.

If you can’t answer how your product/service does one or more of these things, your job isn’t done.

AI can be used to create personalized songs that you can then gift to people you care about. We call that entertainment!

Reliable AI chatbots empower users by delivering relevant help docs so that they don’t have to wait in long queues — an excellent way to educate users and help them help themselves.

Vector search [9] allows users to find search results based not just on how well their keywords match but on the meaning behind them. Better search results, faster? Problem solved.

An AI Reset: Moving Forward With Renewed Energy

AI fatigue doesn’t have to be permanent, but we do need to shift our approach.

By this point, we’ve got at least a basic understanding of just how powerful AI is and what it’s capable of. We’ve tested it and applied it in countless ways. Some have been miraculous, and others have been disastrous.

Next, we iterate!

Using AI at the right time, under the right circumstances, and always for the betterment of users — consider this your north star, and you’ll never go wrong.

References

1. https://www.zippia.com/answers/is-technology-growing-exponentially/
2. https://leftronic.com/blog/how-fast-is-technology-growing-statistics
3. https://www.gartner.com/en/documents/5505695
4. https://futurism.com/cnet-for-sale-ai
5. https://www.charlotteobserver.com/news/business/article290791529.html
6. https://www.axios.com/2024/08/06/cnet-ziff-davis-red-ventures#
7. https://www.linkedin.com/in/apoorvajoshi95/
8. https://mdb.link/towards-ai-dc
9. https://mdb.link/vector-search-towards-ai

Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming a sponsor.

Published via Towards AI