{"slug": "what-does-it-mean-to-know", "title": "What Does It Mean to Know?", "summary": "A developer explores the nature of knowledge and intelligence by questioning what it means for a machine to 'know' something. They argue that intelligence is not about storing isolated facts or rigid rules, but about understanding degrees of certainty, situational rule application, and deciding what information is worth remembering based on expected future utility.", "body_md": "This is non-standard and first-principal take on AI. I will try and avoid jargon and stick to intuition.\n\nLet's ask much simpler and very old question\n\nWhat does it mean to \"know\" something?\n\nI am not looking for a philosophical answer. Let's say we simply ask\n\nHow does a calculator \"know\" 2 + 2 is 4 not 5?\n\nCalculator does not need to know, It follows some rigid rules.\n\nNow suppose I ask:\n\nHow does some recommendation engine recommend a movie or a book to you?\n\nRules and lots of data.\n\nNow suppose I ask:\n\nWhich one the largest planet in solar system?\n\nThe last one seems different, doesn't it?\n\nWhich one is answered based on knowledge? All of them? last one? or none?\n\nLet's build our very own intelligent machine. Let's not worry about how to build it, rather let's just focus only on what capabilities it will have.\n\nFor example, should it be able to:\n\nThere is no right answer.\n\nLet's see if I decide to choose just three\n\nLet's stop and recheck the choices if I was onto something..\n\nBut,,, **What is a fact?**\n\nFor example,\n\nThese are facts.\n\nNow consider\n\nFacts?\n\nWell... mostly\n\nSome cats have three\n\nSome are born without one\n\nThe certainty has become fuzzy\n\nNow consider:\n\nNow consider:\n\nCertainly not a fact..\n\nSo our intelligent machine can't store facts, isolated facts. It needs to understand\n\ndegreesof certainty and generalizations.\n\nSecond capability in the list, \"Follow rules\" is equally uncertain too. Rules change as per the situations.\n\nYour intelligent car will follow rules like\n\nDrive on the Left or Drive on the Right (within a region it typically is fixed)\n\nBut\n\nShould I overtake now?\n\nleads to rules that are purely situational.\n\nRules aren't changing,\n\napplicationof the rule is.\n\nIf that is countered as\n\nWho decides what is important?\n\nLet's pick following facts\n\nTomorrow...\n\nWhich one of these facts our intelligent system must remember?\n\nNot all of them.\n\nWhy?\n\nBecause intelligence isn't just remembering. It is about **what deserves to be remembered**.\n\nThat is much harder than it seems.\n\nStill circling around fact..\n\nSuppose we built two systems\n\n**System A** remembers everything forever.\n\n**System B** forgets almost everything, keeping only what seems likely to matter later.\n\nAt first glance, System A sounds smarter.\n\nIt isn't.\n\nRemembering everything eventually becomes a burden.\n\nThe real challenge is deciding what can be forgotten safely.\n\nSystem A will become impractical soon while System B is weirdly difficult. For example, from the facts that we learned, **My name is X**, **I am allergic to Peanuts** seem important enough to remember tomorrow, they are not changing. Where my car is parked, may not be, car may have been moved for various unknown reasons and subsequent facts about the car and the parking lot itself.\n\nThat single statement alone holds a significant idea that appears over and over again in AI, databases, and even in human cognition.\n\nInformation has no intrinsic importance. Its importance depends on future use\n\nLet me elaborate with an example.\n\n\"My name is X\"\n\nIt can be safely remembered as it does not change. Tomorrow the name is going to be X. But let's suppose the name is name of a waiter, you are talking to, in some cafe, you rarely frequent. Does this waiter or you need to remember each other's names?\n\nProbably NO.\n\nNow suppose the same setting but not a waiter, your family doctor.\n\nShould your family doctor remember your name?\n\nAbsolutely. Same fact, different consumer, different retention needs.\n\nNow about the Peanut allergy, that seemed important anyday, not just tomorrow, But why?\n\nIt is not because it is permanent. It's because, **forgetting it has a very high cost**\n\nImportance isn't about permanence. It's about **expected future utility**\n\nNow we are at a very important juncture where our intelligent systems is presented with just two facts.\n\nOne day X asks\n\n\"Can I eat Food Y?\"\n\nNobody explicitly taught a new fact to the intelligent system that \"X cannot eat food Y\". yet it is expected from the would be intelligent system that it would advise X against Food Y.\n\n**Questions:**\n\nWhere did that new knowledge came from?\n\nIt wasn't part of the training set.\n\nIt isn't a learned fact.\n\nIt wasn't stored in the system.\n\nEssentially the **questions** itself is where the intelligence really begins. And, that's where modern AI starts to diverge from the traditional programming. The crux of the understanding is answer to the question\n\nHow does knowledge become connected?\n\nThat is exactly I will try to chase in this series.", "url": "https://wpnews.pro/news/what-does-it-mean-to-know", "canonical_source": "https://dev.to/justcatdev/what-does-it-mean-to-know-4612", "published_at": "2026-07-18 20:15:38+00:00", "updated_at": "2026-07-18 20:27:58.329091+00:00", "lang": "en", "topics": ["artificial-intelligence"], "entities": [], "alternates": {"html": "https://wpnews.pro/news/what-does-it-mean-to-know", "markdown": "https://wpnews.pro/news/what-does-it-mean-to-know.md", "text": "https://wpnews.pro/news/what-does-it-mean-to-know.txt", "jsonld": "https://wpnews.pro/news/what-does-it-mean-to-know.jsonld"}}