A recent LinkedIn post by Professor Rose Luckin questions whether schools should commit important resources to generate artificial intelligence (AI).
Luckin, a professor at the Institute of Education, University of London and founder of Educate Ventures Research Limited, is an international authority in education. She analysed the latest major technology revenues and impact on schools, saying, “This week's massive technology revenue revealed incredible truths. Companies spent over $350 billion on AI infrastructure in 2025, but are unable to effectively demonstrate meaningful revenue from generated AI.”
Big technology spending patterns
Luckin said: “We plan to invest more than $100 billion in 2026. From 50% to 50% of our expected revenue. Microsoft will pour $120 billion into our data centers. Amazon will not be able to meet AI demand, but AWS' growth is “unfortunate.” Even Zuckerberg acknowledges that Meta “doesn't expect generative AI to have a significant impact on revenues for at least the next few years.” ”
She also highlights the link between performance and traditional sales growth rather than AI features. “NVIDIA is leading performance at +1,247% because it sells infrastructure. Apple shows +45% with minimal AI investment.
Traditional AI-paired AI
In her post, Luckin encourages readers to “distinguish between AI (pattern recognition, data analysis, recommended systems) and generator AI (content creation, conversation interface). She said, “Traditional AI is already working. It drives meta ad improvements and Google's search algorithms. Generation AI is a $350 billion bet with unclear returns.”
Impact on education
Turning to the schools, Luckin admits, “Schools are caught up in generative AI hype, but the biggest tech companies still can't even explain revenue models, and even explain exactly how basic models work!” She acknowledges that “UK DFEs have done a great job with the expectations of generative AI safety, the required measurement approach,” but that they “can't fix the underlying fundamental models issues discussed last week.”
She raises the question, “If Meta describes their investment as a 'wonderful shot in the dark', then cash-bound schools should make a similar bet.”
Educational Value and Measured Adoption
Luckin derives the distinction between traditional AI's “clear educational applications” – the “mainly theoretical” advantages of generated AI, such as personalized learning paths, automated markings, and data analysis.
She writes:
Her conclusion is, “Schools should learn from Big Tech's approach, rather than copying it. They use proven AI applications to promise that even multi-billion dollar businesses can't justify yet. Educational interests are different from Silicon Valley.
