AI Hype Falters as 2.2 Million GNoME Materials and Khanmigo Tutor Fail to Deliver
Updated
Updated · The Conversation · May 27
AI Hype Falters as 2.2 Million GNoME Materials and Khanmigo Tutor Fail to Deliver
3 articles · Updated · The Conversation · May 27
Khan Academy’s Khanmigo and Google DeepMind’s GNoME are cited as marquee AI projects that fell short of sweeping promises, with Khanmigo called “a non-event” and GNoME’s 2.2 million materials later challenged by experts.
Khanmigo was built to transform education through AI tutoring, but students largely did not engage, and universities still lack hard evidence that AI tutors improve learning outcomes.
Workplace AI agents have also underperformed: startup Every found AI line-manager agents produced endless chatter and no decisions, while token costs ran into the millions.
That gap between promise and results is widening as OpenAI and Anthropic leaders soften earlier warnings of entry-level job wipeouts and companies such as Uber and Microsoft report AI budgets exceeding human labor costs.
The article argues these “nothingburgers” will keep surfacing as adoption rises, making AI literacy crucial for separating genuine breakthroughs from industry hype.
With AI 'slop' and failed promises now common, is the AI revolution already over?
From 'job apocalypse' to 'augmentation,' what is the true impact of AI on our careers?