10h agoGary Marcus argues LLMs will never achieve profitability, predicting a shift toward highly efficient neurosymbolic AIG1 top authorHe warns that current massive AI investments are premature.Original postGAGary Marcus@GaryMarcusThree Predictions: 1. Some form of AI, probably neurosymbolic in nature, will come that is far more economical and data- and energy-efficient than LLMs, and it will make an absolute fortune. 2. LLMs, on the other hand, will never be all that profitable (aside from the chip companies selling shovels in the gold rush). 3. Today’s gigantic bets are premature, and most won’t pay off.10:43 PM · Jun 3, 2026SentimentSentimentPos66.7%Neg33.3%Positive users agree with Gary Marcus that neurosymbolic AI will outperform LLMs in efficiency and profits, while negative users dismiss his predictions as consistently wrong.13 comments with sentiment. View comments.3 more postsGAGary Marcus@GaryMarcus·2hReply@IntuitMachine huge and neurosymbolic is what i am saying.View onGAGary Marcus@GaryMarcus·2hReply@IntuitMachine semantically coherent is certainly a prerequisite it trustworthy AI, and unlocking the trillions.
GAGary Marcus@GaryMarcus·2hReply@IntuitMachine semantically coherent is certainly a prerequisite it trustworthy AI, and unlocking the trillions.
GAGary Marcus@GaryMarcus·2hReply@IntuitMachine and that harness is built out of the kinds of symbolic primitives that I advocated for decades, throughout my career.View on
