How this term is usednnWhen the channel says ultimate compression, it is describing why invoking an influential thinker by name is such a powerful prompting technique. As the host puts it:nn> **GPT is magical when it comes to names. Names are like the ultimate compression.**nnA name is maximally compressed because a single proper noun subsumes an entire body of work. The video demonstrates this with the New York Times case study:nn- Clayton Christensen — two words that pull in disruption theory and Jobs to Be Done, surfacing the risk of nimbler, cheaper competitors gradually moving upmarket. n- Michael Porter — summons the Five Forces framework, revealing low barriers to digital entrants, weakening willingness to pay, substitutes from citizen journalism, and the growing bargaining power of star journalists. n- Ben Thompson — brings aggregation theory, highlighting the threat of platforms like Google and Facebook aggregating the Times' own content. nnEach name delivers a whole intellectual toolkit in a few words. Once the frameworks are loaded, the model builds a tapestry of knowledge — you can watch three thinkers converse: see where they agree and follow the fault lines where they disagree.nnnnContrast with the broad-prompt failurennThe term only makes sense against the alternative. A generic prompt like critique this business model gets no names and no constraints, so the model defaults to web-searched boilerplate — the host calls the result buzzword bingo. The compression technique works because it gives the model shape and direction: you must constrain it, otherwise it won't give you a good example. The same principle applies to follow-ups — compress further by asking for one named strategy per thinker, each no longer than a paragraph.nnnnnn- Ultimate compression means a person's name encodes their entire framework, so naming a thinker loads their ideas into the model in almost no words. n- Demonstrated with Christensen (disruption), Porter (Five Forces), and Thompson (aggregation theory) analyzing the New York Times' ad-to-subscription shift. n- Names beat topic descriptions because they summon specific, connected thinking instead of consensus mush. n- Use it by telling the model which thinkers to draw on — and telling it not to search the web first. n