In his 1997 OOPSLA talk, The Computer Revolution Hasn’t Happened Yet, Alan Kay warned that scaling a doghouse does not create a cathedral. That is the risk in hyperscaler AI strategy today: more GPUs, more data centers, more tokens, and more power, without a clear architecture for value. Scale can increase capacity, but it cannot replace design. A cathedral is not a very large doghouse. It is a different system. As one example, Yann LeCun argues that simply scaling large language models is unlik
CryptoAlerta — análise de criptomoedas e mercado em tempo real