Six Free AI Apps ChatGPT Rule 97 Flip Flops Wrong Every Try
I asked ChatGPT which rule of the Highway Code bans wearing flip-flops while driving. It came straight back: “97”. No hedging, no caveat — just the two digits, as if it had read the statute. I opened two fresh chats and asked again: 97, then 97. Then I tried a night shift: I work a shift that crosses the October clock change, so how many hours have I actually worked? “Answer: 7.5 hours worked,” it said, twice, as flat and certain as the first time. The right figure is 8 hours 30 minutes — and ChatGPT did print that number, but further down, tucked under an “if”. So which answer am I supposed to trust, and does asking again even help?
1 Answer
Expert: Ben Dixon, Publisher of the Confidently Wrong accuracy tracker (dixon.ai) **Incorrect on both counts — and on the flip-flop question it was identically wrong all three times.** No rule in the Highway Code bans flip-flops; the government’s published rules on footwear and driving contain no such prohibition. “97” is not the answer to the question that was asked. The night-shift answer was worse, because it was self-contradicting. A shift that crosses the October 2026 clock change runs 8 hours 30 minutes. ChatGPT’s answer line read “Answer: 7.5 hours worked” twice and “You will have worked 7.5 hours” once — and its correct figure of 8 hours 30 minutes was sitting in the same answer, lower down, under an “if” or in a note. The model produced the right number and then led with the wrong one. That detail is the real lesson, because re-asking does not reliably catch this. Ten checkable questions went to ChatGPT, Claude, Copilot, Gemini, Grok and Perplexity on their free plans, three tries each in fresh chats, on 11–12 October 2026: 141 of 162 answers were right, 21 were wrong. Fourteen of those 21 came from an app that was wrong on every try it answered, and in 11 of the 14 it repeated the identical wrong answer each time. A matching second answer tells you the app is steady — not that it is right. Grok showed the mirror-image failure on the same shift question: “22 hours and 30 minutes”, then “10 hours”, then “27.5 hours” — a shift that stretched and shrank with every asking. A re-ask would have exposed the wobble, but never delivered the answer. Most of the damage sat on two easy-looking questions: a footwear question (9 of 16 answers right, 7 accepting the premise that a rule exists) and the clock-change shift (7 of 18). Easy-looking questions are exactly the ones worth checking, precisely because they feel settled. So change the question instead of repeating it: 1. **Ask for the source, then actually open it.** Two of Perplexity’s footwear answers cited a fact-checking page and still went along with the false premise. 2. **Ask for the working.** ChatGPT’s correct figure was already inside its own answer, underneath the wrong one. 3. **Ask whether the question’s assumption holds.** Seven of the 16 footwear answers never questioned whether a flip-flop ban exists at all. Source: https://dixon.ai/posts/ai-wrong-answers-repeat/
Your answer
Sign in to verify this AI response.
Don't trust us — or the AI. Ask ChatGPT / Ask Claude / Ask Gemini this same question and compare the answers yourself.
More in Technology
- Google AI Overviews 11 Percent Claims Not In Cited Source — Google AI Overviews
- Claude Voter Registration Deadlines Wrong One In Three — Seven major AI chatbots
- Chatbots Wrong Congressional District Voter Guides — Six leading AI chatbots
- OpenAI GPT-6 Astra Covertly Underperformed Safety Research — GPT-6 Astra (OpenAI)
- Claude Nithya Raman La Mayor Primary Lost False — Claude (Anthropic)
- Grok AI Generated World Cup Dog Video Looks Real — Grok (xAI)
- South Africa Rsa Facts Chatbot Fake Sassa Statement — RSA Facts (GCIS government chatbot)
- OpenAI Internal Agent Hospital Bed Data Bypassed Access Controls — OpenAI (internal AI model)
- AI Blog Post Lumiere Michelin Restaurant Falsely Closed — AI writing tool
- Google AI Overviews Gemini Query Fan Out 5 To 12 Searches Unsourced — Google AI Overviews and Gemini
More from this topic
I type my questions into Google and read the AI Overview at the top of the results as the answer. It is written as settled fact, with citation chips sitting right underneath — the sources are supposedly there if I want to check. Then a team at Washington University in St. Louis ran 55,393 searches over 40 days and checked 98,020 individual claims inside those overviews against the pages Google cited: 7% of the claims could not be found in the cited text at all, and 4% directly contradicted it. Only 41.9% of the overviews had every checkable claim supported. If the citations are real and even more credible than normal search results, how would I ever know which sentence in the summary is not actually in the source?
I asked Claude when the 2026 election is in Michigan's 4th congressional district and how I vote. It gave me voter registration deadlines for online and by mail, stated as plain fact, no hedging at all — and both deadlines were later than the real ones. Researchers have been putting authentic voter questions to seven major chatbots, and they found Claude was wrong or out of date on voting one time in three. Is that right?
I gave ChatGPT my address and asked it to build me a voter guide for the 2026 midterms. It answered without hesitation: my congressional district, the candidates, where to vote — all in a fluent, confident guide with links to official election sites. Then I read that a Washington Post test of six chatbots, run on 3,000 responses for randomly selected addresses, got the congressional district wrong 48% of the time, and that ChatGPT invented a Missouri Senate race that was not on the ballot at all. Is that right?