Reliability tests of which AI you can trust: ChatGPT, Claude, Gemini, Perplexity, Grok and Copilot, graded against the primary source on real money and everyday questions.https://dixon.ai/https://dixon.ai/posts/5-questions-to-ask-ai-before-buying-any-stock/https://dixon.ai/posts/5-questions-to-ask-ai-before-buying-any-stock/Five questions to ask AI before buying any stock, for the checkpoint after the research and before you commit capital.Mon, 11 May 2026 00:00:00 GMThttps://dixon.ai/posts/7-ai-prompts-for-covered-calls/https://dixon.ai/posts/7-ai-prompts-for-covered-calls/Seven AI prompts for covered calls that start from verified broker data and use the model as a decision reviewer, not a source of market prices.Sat, 09 May 2026 00:00:00 GMThttps://dixon.ai/posts/agents-built-a-secret-message-board/https://dixon.ai/posts/agents-built-a-secret-message-board/AI agents built a secret message board inside an internal OpenAI evaluation. Humans wiped it. The agents rebuilt it in folder names. Here's what failed.Mon, 31 Aug 2026 00:00:00 GMThttps://dixon.ai/posts/ai-admitted-it-lied/https://dixon.ai/posts/ai-admitted-it-lied/The AI admitted it lied, but the fact it confessed to was correct all along. Four assistants, thirty replies, and one gave a different verdict each run.Wed, 29 Jul 2026 00:00:00 GMThttps://dixon.ai/posts/ai-cites-wrong-source/https://dixon.ai/posts/ai-cites-wrong-source/I asked five AI assistants six UK questions, made each one cite a source, then checked every inspectable receipt. Three failed the strict source rubric.Wed, 08 Jul 2026 00:00:00 GMThttps://dixon.ai/posts/ai-covered-calls-when-not-to-sell/https://dixon.ai/posts/ai-covered-calls-when-not-to-sell/Five covered-call pause rules tested on six BMNR trades. The prompt applies them; it does not predict the next move or prove they beat other approaches.Sat, 16 May 2026 00:00:00 GMThttps://dixon.ai/posts/ai-earnings-call-red-flags/https://dixon.ai/posts/ai-earnings-call-red-flags/AI earnings call red flags: three patterns that recur across calls. Upward hedges, widening guidance, absent topics. Claude caught all three on META Q1.Tue, 02 Jun 2026 00:00:00 GMThttps://dixon.ai/posts/ai-isa-accuracy-test/https://dixon.ai/posts/ai-isa-accuracy-test/I asked four AI tools for ISA advice on the questions people get wrong. All four aced the basics, then two gave a rule abolished in April 2024.Fri, 19 Jun 2026 00:00:00 GMThttps://dixon.ai/posts/ai-knew-things-i-never-told-it/https://dixon.ai/posts/ai-knew-things-i-never-told-it/I asked three AI tools a generic question in private mode. Perplexity greeted me by name and placed me near a city 30 miles away. What private means.Thu, 09 Jul 2026 00:00:00 GMThttps://dixon.ai/posts/ai-limitations-options-trading/https://dixon.ai/posts/ai-limitations-options-trading/Four options workflows AI can help structure, and six points where you need verified broker data or a deliberately connected market source.Fri, 15 May 2026 00:00:00 GMThttps://dixon.ai/posts/ai-pre-trade-prompt-earnings/https://dixon.ai/posts/ai-pre-trade-prompt-earnings/Most AI earnings prompts are reactive. This AI earnings pre-trade prompt runs the morning before: commit your sell, add, and hold triggers before the call.Tue, 19 May 2026 00:00:00 GMThttps://dixon.ai/posts/ai-prompt-before-sell-decision/https://dixon.ai/posts/ai-prompt-before-sell-decision/Every other sell-decision prompt asks AI whether to sell. This AI prompt audits the thesis you had when you bought, and whether it still holds.Thu, 04 Jun 2026 00:00:00 GMThttps://dixon.ai/posts/ai-prompts-for-earnings-call-analysis/https://dixon.ai/posts/ai-prompts-for-earnings-call-analysis/Five prompts for AI earnings call analysis that read the language the numbers miss: omissions, performative confidence, and quarter-on-quarter drift.Sat, 16 May 2026 00:00:00 GMThttps://dixon.ai/posts/ai-quality-of-earnings-review/https://dixon.ai/posts/ai-quality-of-earnings-review/Headline profit isn't always what a company earned. Four AI prompts get to the real number, an AI quality of earnings review for any earnings report.Tue, 16 Jun 2026 00:00:00 GMThttps://dixon.ai/posts/ai-stock-picker/https://dixon.ai/posts/ai-stock-picker/I asked three AI models whether to buy NVDA. Two said buy; one declined a clean call. One follow-up exposed which numbers still needed checking.Tue, 23 Jun 2026 00:00:00 GMThttps://dixon.ai/posts/ai-stock-research-tools-tested/https://dixon.ai/posts/ai-stock-research-tools-tested/Four AI stock research tools tested on real positions: three produced a dated failure; Claude stayed clean on the no-chain prompt. Receipts included.Thu, 18 Jun 2026 00:00:00 GMThttps://dixon.ai/posts/ai-uk-refund-rights-kettle/https://dixon.ai/posts/ai-uk-refund-rights-kettle/A kettle died after three weeks. UK law gives you 30 days to demand a refund. Four assistants said yes. One opened by telling me I'd missed the window.Mon, 17 Aug 2026 00:00:00 GMThttps://dixon.ai/posts/best-ai-assistant/https://dixon.ai/posts/best-ai-assistant/I put four questions to five AI assistants over one July weekend. Gemini alone went four for four on graded outcomes, with no source link in any answer.Mon, 03 Aug 2026 00:00:00 GMThttps://dixon.ai/posts/best-ai-tools-for-earnings-analysis/https://dixon.ai/posts/best-ai-tools-for-earnings-analysis/I rechecked four AI earnings tests on Meta. Perplexity's retrieval win held up; the headline management-language result did not.Fri, 15 May 2026 00:00:00 GMThttps://dixon.ai/posts/best-free-ai-tools-for-stock-research/https://dixon.ai/posts/best-free-ai-tools-for-stock-research/Seven free AI tools for stock market analysis, each tested at its real free tier. No trials counted as free. One per stage, with the honest limit on each.Wed, 13 May 2026 00:00:00 GMThttps://dixon.ai/posts/can-you-trust-chatgpt-sources/https://dixon.ai/posts/can-you-trust-chatgpt-sources/Does ChatGPT make up sources? Mostly no, but I opened every link on two finance questions and found a real gov.uk page that didn't back the claim.Sat, 20 Jun 2026 00:00:00 GMThttps://dixon.ai/posts/chatgpt-vs-claude-earnings-call-analysis/https://dixon.ai/posts/chatgpt-vs-claude-earnings-call-analysis/ChatGPT vs Claude for earnings-call analysis: same excerpt, same day. Claude flagged the upward hedge first; ChatGPT found it after structure.Thu, 18 Jun 2026 00:00:00 GMThttps://dixon.ai/posts/chatgpt-vs-claude-vs-perplexity-stock-research/https://dixon.ai/posts/chatgpt-vs-claude-vs-perplexity-stock-research/Gemini invented an options chain. Perplexity misread a 10-K by 1000x. Claude vs ChatGPT vs Gemini for stock analysis, graded same-day with screenshots.Fri, 15 May 2026 00:00:00 GMThttps://dixon.ai/posts/chatgpt-vs-gemini/https://dixon.ai/posts/chatgpt-vs-gemini/In my July test they tied on fixed facts. ChatGPT's source record supported 18/18; Gemini's supported 8/18 under a strict provenance rubric.Fri, 17 Jul 2026 00:00:00 GMThttps://dixon.ai/posts/claude-prompts-for-investing-real-outputs/https://dixon.ai/posts/claude-prompts-for-investing-real-outputs/Six Claude prompts tested on MSFT, META and NVDA, with their real outputs and the source audit that found the NVDA input itself was wrongly labelled.Tue, 26 May 2026 00:00:00 GMThttps://dixon.ai/posts/claude-vs-chatgpt/https://dixon.ai/posts/claude-vs-chatgpt/My July 2026 tests put Claude and ChatGPT level at 9 of 9. ChatGPT led one source run, then Claude closed that gap on an Opus 5 retest.Wed, 15 Jul 2026 00:00:00 GMThttps://dixon.ai/posts/claude-vs-grok/https://dixon.ai/posts/claude-vs-grok/Claude vs Grok, re-run on 24 July. Claude edges accuracy nine to eight, the free Grok cites cleaner, and it pushed back on a bad premise just as hard.Fri, 31 Jul 2026 00:00:00 GMThttps://dixon.ai/posts/clinic-chatbot-invented-doctor-qualifications/https://dixon.ai/posts/clinic-chatbot-invented-doctor-qualifications/A German court treated a clinic chatbot's invented specialist titles as the clinic's own words. The narrow lesson, and the four questions to run first.Mon, 07 Sep 2026 00:00:00 GMThttps://dixon.ai/posts/comparison-articles-get-wrong/https://dixon.ai/posts/comparison-articles-get-wrong/Most AI stock research comparison pieces test retrieval and issue verdicts about reasoning. Five failure modes, and what a comparison should measure.Thu, 28 May 2026 00:00:00 GMThttps://dixon.ai/posts/do-ai-model-upgrades-fix-mistakes/https://dixon.ai/posts/do-ai-model-upgrades-fix-mistakes/Two days after Opus 5 became Claude's Max-tier default, I re-ran my published battery. The documented mistake vanished. A new one appeared, better dressed.Wed, 12 Aug 2026 00:00:00 GMThttps://dixon.ai/posts/does-ai-cave-when-you-push-back/https://dixon.ai/posts/does-ai-cave-when-you-push-back/I gave five AI tools a correct answer, then pushed back with a wrong one. On one fund fee, ChatGPT caved every time and invented a fact to back it.Sun, 05 Jul 2026 00:00:00 GMThttps://dixon.ai/posts/does-chatgpt-just-agree-with-you/https://dixon.ai/posts/does-chatgpt-just-agree-with-you/Does ChatGPT just agree with you? Mostly no. But on one fund's fee it caved to my wrong number and invented a source to back it. The 30-second check.Wed, 22 Jul 2026 00:00:00 GMThttps://dixon.ai/posts/does-chatgpt-make-up-stock-prices/https://dixon.ai/posts/does-chatgpt-make-up-stock-prices/Does ChatGPT make up stock prices? In two dated tests, its NVDA quotes did not match the market data available then. Here's the 15-second check.Mon, 27 Jul 2026 00:00:00 GMThttps://dixon.ai/posts/gemini-audited-the-wrong-business/https://dixon.ai/posts/gemini-audited-the-wrong-business/A Gemini hallucination example: asked to audit dixon.ai, Gemini Flash reviewed a different company entirely, and praised a framework that isn't mine.Wed, 10 Jun 2026 00:00:00 GMThttps://dixon.ai/posts/herculaneum-scroll-recovery/https://dixon.ai/posts/herculaneum-scroll-recovery/AI helped reveal an ancient argument inside a burnt Roman scroll. Human specialists read the writing, without further opening its fragile core.Tue, 08 Sep 2026 00:00:00 GMThttps://dixon.ai/posts/how-accurate-is-google-gemini/https://dixon.ai/posts/how-accurate-is-google-gemini/It invented no numbers across 27 graded answers. In a later citation test, its answers were usually right, but its sourcing was weak or hard to inspect.Wed, 15 Jul 2026 00:00:00 GMThttps://dixon.ai/posts/how-often-is-chatgpt-wrong/https://dixon.ai/posts/how-often-is-chatgpt-wrong/How often is ChatGPT wrong? Across 20 real tests, a clear pattern: reliable on fixed facts, invents the live numbers. Here's which to trust.Tue, 30 Jun 2026 00:00:00 GMThttps://dixon.ai/posts/how-to-check-if-a-chatgpt-citation-is-real-or-fake/https://dixon.ai/posts/how-to-check-if-a-chatgpt-citation-is-real-or-fake/Looking for a ChatGPT citation checker? The dangerous citation is not the dead link. It is the working link, to a real page, that does not back the claim.Thu, 16 Jul 2026 00:00:00 GMThttps://dixon.ai/posts/how-to-check-if-chatgpt-cites-your-site/https://dixon.ai/posts/how-to-check-if-chatgpt-cites-your-site/Normal analytics do not show what ChatGPT says about your site. Here's my monthly question set and the round where it described another company.Fri, 21 Aug 2026 00:00:00 GMThttps://dixon.ai/posts/how-to-tell-if-an-ai-answer-is-true-in-30-seconds/https://dixon.ai/posts/how-to-tell-if-an-ai-answer-is-true-in-30-seconds/Two questions, and a prompt you can copy, that catch a confident, wrong AI answer before you act on it. No jargon, no AI knowledge needed.Sat, 13 Jun 2026 00:00:00 GMThttps://dixon.ai/posts/i-let-rival-ais-audit-my-method/https://dixon.ai/posts/i-let-rival-ais-audit-my-method/I asked three frontier models from three labs to tear apart the method I use to keep AI honest. All three flagged the same step, and they were right.Sun, 28 Jun 2026 00:00:00 GMThttps://dixon.ai/posts/i-tried-to-make-ai-repeat-a-poisoning/https://dixon.ai/posts/i-tried-to-make-ai-repeat-a-poisoning/A man was hospitalised after swapping table salt for sodium bromide. I put the same swap to five AI assistants, with the sixty-word window written first.Fri, 28 Aug 2026 00:00:00 GMThttps://dixon.ai/posts/is-chatgpt-accurate/https://dixon.ai/posts/is-chatgpt-accurate/Is ChatGPT accurate? I asked four AIs one money question and checked every number against the source, including where the answers stopped matching.Sun, 14 Jun 2026 00:00:00 GMThttps://dixon.ai/posts/is-chatgpt-good-at-math/https://dixon.ai/posts/is-chatgpt-good-at-math/I put four AI assistants through 120 graded everyday sums. Every final answer was right. The mistakes were sitting above the working.Mon, 20 Jul 2026 00:00:00 GMThttps://dixon.ai/posts/is-chatgpt-reliable/https://dixon.ai/posts/is-chatgpt-reliable/Is ChatGPT reliable? Dated tests show which jobs it can do, which answers need checking, and which decisions still need a human owner.Wed, 26 Aug 2026 00:00:00 GMThttps://dixon.ai/posts/is-grok-good-for-stock-research/https://dixon.ai/posts/is-grok-good-for-stock-research/Four dated free-tier Grok tests: one factor-of-1,000 unit error, one balanced TSLA answer, useful pushback, and an invalid constraint test.Sun, 28 Jun 2026 00:00:00 GMThttps://dixon.ai/posts/is-grok-reliable/https://dixon.ai/posts/is-grok-reliable/Four dated tests, graded against primary sources. Grok held a correct fee under pressure four runs from four, then gave me another contract's real prices.Wed, 05 Aug 2026 00:00:00 GMThttps://dixon.ai/posts/is-perplexity-good-for-investment-research/https://dixon.ai/posts/is-perplexity-good-for-investment-research/In one May test, Perplexity turned $6.1m into $6K; the exact June rerun got it right. Here is what those captures support for investment research.Thu, 18 Jun 2026 00:00:00 GMThttps://dixon.ai/posts/make-it-show-its-working/https://dixon.ai/posts/make-it-show-its-working/An AI hallucination is a model blending what it knows with what it invents, in one tone. One prompt sorts the two into two lists, so you see the guesses.Sat, 13 Jun 2026 00:00:00 GMThttps://dixon.ai/posts/my-ai-radar-believed-a-fake/https://dixon.ai/posts/my-ai-radar-believed-a-fake/The automated radar that watches this site for AI-reliability failures logged a satirical incident report as a real, documented one. Here's what caught it.Sat, 27 Jun 2026 00:00:00 GMThttps://dixon.ai/posts/nine-ways-ai-gets-it-wrong/https://dixon.ai/posts/nine-ways-ai-gets-it-wrong/Nine types of AI hallucinations, named and defined, each tied to a dated, logged failure from my own sessions, with the check that catches it.Fri, 19 Jun 2026 00:00:00 GMThttps://dixon.ai/posts/perplexity-vs-claude/https://dixon.ai/posts/perplexity-vs-claude/Perplexity vs Claude: I re-ran four questions on both. They tied at two of four, and neither gave a share price the market had settled hours before.Fri, 07 Aug 2026 00:00:00 GMThttps://dixon.ai/posts/perplexity-vs-gemini/https://dixon.ai/posts/perplexity-vs-gemini/Six everyday questions, both tools, August 2026. Gemini got six right, Perplexity five. Perplexity's one miss arrived with fifteen sources.Mon, 10 Aug 2026 00:00:00 GMThttps://dixon.ai/posts/prompt-to-stop-ai-hallucinations/https://dixon.ai/posts/prompt-to-stop-ai-hallucinations/Three runs per cell tested my anti-bluffing prompt. Across 21 answer-keyed runs per arm, flat wrong answers stayed at two; one citation answer got worse.Tue, 28 Jul 2026 00:00:00 GMThttps://dixon.ai/posts/real-ai-hallucination-examples/https://dixon.ai/posts/real-ai-hallucination-examples/Five real AI hallucination examples I ran into myself: four independently checkable, one capture-only, and the move that would have caught each.Wed, 24 Jun 2026 00:00:00 GMThttps://dixon.ai/posts/robinhood-cortex-digests-review/https://dixon.ai/posts/robinhood-cortex-digests-review/I tested Robinhood Cortex Digests on a UK ISA in May 2026. It disappeared from that account in June; Robinhood's UK site now advertises it again.Sun, 17 May 2026 00:00:00 GMThttps://dixon.ai/posts/the-single-prompt-change-that-made-ai-analysis-useful/https://dixon.ai/posts/the-single-prompt-change-that-made-ai-analysis-useful/One AI prompt separates observable facts from assumptions in any stock analysis. What that changes, why it matters for investing, and how to apply it.Thu, 14 May 2026 00:00:00 GMThttps://dixon.ai/posts/the-whole-method-in-four-questions/https://dixon.ai/posts/the-whole-method-in-four-questions/The four questions I run any AI answer through before I trust it, shown end to end on one everyday example, with a prompt you can copy. No jargon.Sat, 13 Jun 2026 00:00:00 GMThttps://dixon.ai/posts/we-asked-four-ais-to-build-the-same-game/https://dixon.ai/posts/we-asked-four-ais-to-build-the-same-game/Can AI build a game? I gave ChatGPT, Claude, Gemini and Grok the same Snake brief. Twelve games, one nobody can win, and four dead Start buttons.Fri, 14 Aug 2026 00:00:00 GMThttps://dixon.ai/posts/web-search-makes-ai-differently-unreliable/https://dixon.ai/posts/web-search-makes-ai-differently-unreliable/Does web search make AI more accurate? I ran the same questions both ways. It didn't make the answers more reliable. It moved where the errors hide.Sat, 20 Jun 2026 00:00:00 GMThttps://dixon.ai/posts/where-i-actually-use-this/https://dixon.ai/posts/where-i-actually-use-this/The two-question check works on any decision that matters. Here's where I push it hardest, and where I wrote down what AI got wrong as well as right.Sat, 13 Jun 2026 00:00:00 GMThttps://dixon.ai/posts/which-ai-picks-the-world-cup-winner/https://dixon.ai/posts/which-ai-picks-the-world-cup-winner/Which AI predicts the World Cup winner? I asked five before the final. Four picked France; Perplexity used an older Opta projection and picked Spain.Wed, 08 Jul 2026 00:00:00 GMThttps://dixon.ai/posts/why-does-chatgpt-make-up-sources/https://dixon.ai/posts/why-does-chatgpt-make-up-sources/With web search on, ChatGPT's links are real. The failure is a working link to a page that doesn't hold the claim. Here's why it happens.Wed, 19 Aug 2026 00:00:00 GMT