Has Apple exposed the core problem with AI?

Apple has just sent a seismic shockwave through the tech community. Their senior machine learning experts have recently released a paper that peels away the hype to expose the core truths about LLMs. It makes worrying reading…

The AI industry stands accused of exaggerating the capabilities and scaling potential of leading AI models. Claims that the most advanced models can now ‘reason’ are brutally dismissed as an “illusion of thinking”. This report certainly doesn’t mince words: ‘Extensive experimentation across diverse puzzles shows that frontier, large reasoning models face a complete accuracy collapse beyond certain complexities’.

Is Apple’s paper a much-needed, AI reality check? Or is it a cynical attempt to rubbish the game-changing achievements of rivals? Apple has, after all, been widely accused of falling far behind the competition in the AI space.