When did computers first start trying to understand human language - and how spectacularly did they fail?

The answer

Early natural language processing began with ambitious machine translation efforts in the 1950s, driven by Cold War urgency to translate Russian scientific documents. The Georgetown-IBM experiment of 1954 demonstrated automatic translation of 60 Russian sentences into English using just six grammar rules and 250 vocabulary words, leading to wildly optimistic predictions. However, the infamous (though possibly apocryphal) back-translation of 'the spirit is willing but the flesh is weak' into Russian and back as 'the vodka is good but the meat is rotten' illustrated the deep difficulty. The 1966 ALPAC (Automatic Language Processing Advisory Committee) report concluded that machine translation was slower, less accurate, and more expensive than human translation, causing a dramatic reduction in NLP funding that lasted over a decade.