Mijn systeem denkt niet in antwoorden maar in waarschijnlijkhedenDe meeste mensen denken binair: waar of onwaar, relevant of irrelevant. Maar de interessante antwoorden zitten zelden in die uitersten maar in de gradaties ertussen. Iets kan voor zeventig procent waar zijn in de ene context, en voor dertig procent in een andere. Daar bestaat een term voor: fuzzy logic, een formeel wiskundig systeem dat in 1965 werd bedacht door Lotfi Zadeh en dat draait om lidmaatschapsgraden in plaats van harde grenzen.Dat is precies hoe dit systeem werkt. Holmes zegt: het bewijs wijst voor 0.8 naar deze verklaring. Da Vinci zegt: de analogie klopt voor 0.6. Occam zegt: deze factor draagt voor 0.2 bij, schrap hem. De analisten leveren geen ja-of-nee maar gewogen signalen, en de eindredacteuren wegen die signalen tegen elkaar.

Why This Matters: From Probabilistic Guesses to Deterministic AnswersGrounding your AI in a Knowledge Graph delivers three non-negotiable enterprise advantages:1. Accuracy: Answers are derived from an explicit model of your business, not from statistical correlations in text. You eliminate both factual and relational hallucinations.2. Explainability: Every answer comes with a query that shows exactly how it was derived, which entities were connected and which rules were applied. This turns the AI from a black box into a transparent tool.3. Architectural Stability: The semantic layer, the ontology, remains stable even as the underlying systems change. When you migrate your CRM, you simply update the mapping to the ontology. Your AI, analytics, and dashboards continue to work without interruption. This is agility where it counts.

One of my favourite pieces on LLM security this year is The Normalization of Deviance in AI by security researcher Johann Rehberger. Johann describes the “Normalization of Deviance” phenomenon, where repeated exposure to risky behaviour without negative consequences leads people and organizations to accept that risky behaviour as normal. This was originally described by sociologist Diane Vaughan as part of her work to understand the 1986 Space Shuttle Challenger disaster, caused by a faulty O-ring that engineers had known about for years. Plenty of successful launches led NASA culture to stop taking that risk seriously. Johann argues that the longer we get away with running these systems in fundamentally insecure ways, the closer we are getting to a Challenger disaster of our own.

The true and the approximately true are apprehended by the same faculty; it may also be noted that men have a sufficient natural instinct for what is true, and usually do arrive at the truth. Hence the man who makes a good guess at truth is likely to make a good guess at probabilities.

t is evident from what has been said that it is these three subjects, more than any others, about which the orator must be able to have propositions at his command. Now the propositions of Rhetoric are Complete Proofs, Probabilities, and Signs. Every kind of syllogism is composed of propositions, and the enthymeme is a particular kind of syllogism composed of the aforesaid propositions.