Top News

Google's AI Overview feature has become the butt of online jokes after it spectacularly failed a simple letter-counting query. One user on Reddit quipped, "Guess AI can’t replace me for this year."

Google’s AI Overview feature has become the butt of online jokes after it spectacularly failed a simple letter-counting query, declaring there are zero ‘w’s in the word Zimbabwe. The blunder, highlighted in a Reddit post, has users laughing and questioning the reliability of AI-driven search answers and sharing screenshots of equally absurd errors for other letters.

The incident began when a user asked Google, How many ws are there in Zimbabwe? Instead of correctly identifying the single ‘w’ in the seventh position of the eight-letter word, the AI Overview responded that there are zero ‘w’s in the word Zimbabwe. The word has a total of eight letters, but none of them are ‘w’. It even broke it down further, stating: Letter count: The word Zimbabwe contains zero ‘w’s. Total letters: The word Zimbabwe has eight letters in total.

Emboldened by the original error, other Redditors tested similar queries to poke holes in the system. One asked how many ‘e’s are in Zimbabwe, only for the AI to claim there are two, overlooking the single ‘e’ at the end. Another queried the number of a’s, prompting a response of two when there is just one in the fifth spot. These follow-up gaffes amplified the ridicule, with users circulating screenshots that painted Google’s tool as comically inept at basic literacy tasks.

The viral thread drew comparisons to rival AIs, adding fuel to the fire. When one user fed a screenshot of the AI Overview ‘w’ mistake to ChatGPT, it rightly noted there is one ‘w’ in Zimbabwe but stumbled by claiming it appears in the fifth position, a slip that Z-i-m-b-a-b-w-e clearly refutes.

Commenters piled on with sharp wit, "That’s just one way to say that current LLMs have no concept what a word or a letter even is." Another quipped, "Guess AI can’t replace me for this year."

A third vented frustration over deployment choices, "I wish Google didn’t prioritise AI answers even for questions it can’t answer, as in this post. Not that there’s a terribly good reason to be googling this question." And for good measure, someone joked, "Calm down we all know AI can’t count."

This is not the first time Google’s AI Overview has courted controversy. Launched to provide concise, generative summaries atop search results, it has previously suggested eating rocks for nutrition or adding glue to pizza dough, drawing widespread criticism for hallucinations and misinformation. Experts attribute such errors to the probabilistic nature of large language models, which excel at patterns but falter on rote tasks like precise counting without explicit safeguards.