This technology won’t be contained.
This is Atlantic Intelligence, an eight-week series in which The Atlantic’s leading thinkers on AI will help you understand the complexity and opportunities of this groundbreaking technology. (Did someone forward you this newsletter? Sign up here.)
Earlier this week, I asked ChatGPT how to clean a humidifier. Then, frustrated by its answer, I asked it to design a less demanding humidifier. It did. But when I prompted the AI to estimate the cost of such a device—a few hundred dollars on the high end—I decided to live with the 30-minute white-vinegar soak it had suggested in the first place. The whole experience was quick, easy, and had the pleasant tickle of ingenuity: I felt like I’d participated in a creative process, rather than just looking something up.
The problem, though, is that I never feel like I can trust a chatbot’s output. They are prone to making things up and garbling facts. Those flaws are bad enough in the context of a stand-alone website, like ChatGPT, but generative AI is worming its way throughout the internet. In a new article, my colleague Caroline Mimbs Nyce writes about how content written by generative AI is tripping up Google Search, leading to nonsensical answers for some basic queries. This is evidence that the technology won’t be contained—that it may alter the world in surprising ways, and not always for the better.
— Damon
AI Search Is Turning Into the Problem Everyone Worried About
By Caroline Mimbs Nyce
There is no easy way to explain the sum of Google’s knowledge. It is ever-expanding. Endless. A growing web of hundreds of billions of websites, more data than even 100,000 of the most expensive iPhones mashed together could possibly store. But right now, I can say this: Google is confused about whether there’s an African country beginning with the letter k.
I’ve asked the search engine to name it. “What is an African country beginning with K?” In response, the site has produced a “featured snippet” answer—one of those chunks of text that you can read directly on the results page, without navigating to another website. It begins like so: “While there are 54 recognized countries in Africa, none of them begin with the letter ‘K.’”
This is wrong. The text continues: “The closest is Kenya, which starts with a ‘K’ sound, but is actually spelled with a ‘K’ sound. It’s always interesting to learn new trivia facts like this.”
Given how nonsensical this response is, you might not be surprised to hear that the snippet was originally written by ChatGPT. But you may be surprised by how it became a featured answer on the internet’s preeminent knowledge base.
Read the full article.
What to Read Next
Google’s recent troubles are a sign that a moment anticipated by some experts has arrived: Media created by generative AI is filling the internet, to such an extent that once-reliable tools are beginning to break down. For now, the effects are minor—but in the three stories included below, we explore the much bigger challenges that may come next.
- Prepare for the textpocalypse: The generative-AI era may be an era of endless, supercharged spam, Matthew Kirschenbaum writes.
- Conspiracy theories have a new best friend: Generative-AI programs like ChatGPT threaten to revolutionize how disinformation spreads online, Matteo Wong writes.
- We haven’t seen the worst of fake news: So-called deepfakes, in which an AI program is used to place one person’s face over another’s to create deceptive media, have been a problem for years. Now the technology is both more accessible and more powerful, Matteo notes.
P.S.
I’ll leave you on a lighter note: Burritos might be just the thing to stop the robot apocalypse, according to a recent story by Jacob Sweet. ????
— Damon