Archived link: https://archive.ph/Vjl1M
Here’s a nice little distraction from your workday: Head to Google, type in any made-up phrase, add the word “meaning,” and search. Behold! Google’s AI Overviews will not only confirm that your gibberish is a real saying, it will also tell you what it means and how it was derived.
This is genuinely fun, and you can find lots of examples on social media. In the world of AI Overviews, “a loose dog won’t surf” is “a playful way of saying that something is not likely to happen or that something is not going to work out.” The invented phrase “wired is as wired does” is an idiom that means “someone’s behavior or characteristics are a direct result of their inherent nature or ‘wiring,’ much like a computer’s function is determined by its physical connections.”
It all sounds perfectly plausible, delivered with unwavering confidence. Google even provides reference links in some cases, giving the response an added sheen of authority. It’s also wrong, at least in the sense that the overview creates the impression that these are common phrases and not a bunch of random words thrown together. And while it’s silly that AI Overviews thinks “never throw a poodle at a pig” is a proverb with a biblical derivation, it’s also a tidy encapsulation of where generative AI still falls short.
It’s not thinking. It’s just spicy autocomplete; having ingested most of the web, it “knows” that what follows a question about the meaning of a phrase is usually the definition and etymology of that phrase; there aren’t many examples online of anyone asking for the definition of a phrase and being told “that doesn’t exist, it’s not a real thing.” So it does some frequency analysis (actually it’s probably more correct to say that it is frequency analysis) and decides what the most likely words to come after your question are, based on everything it’s been trained on.
But it doesn’t actually know or think anything. It just keeps giving you the next expected word until it meets its parameters.