The Future of Search: Will we still Google it? (www.lrb.co.uk)

Google’s origin story and technical breakthroughs explain both its dominance and why its future is now in question. Larry Page and Sergey Brin solved search relevance with PageRank’s recursive, citation-like weighting of incoming links, and then built massive web crawlers to index the growing web. Facing unreliable cheap hardware, Google invented cluster-wide fault-tolerance and MapReduce (Dean & Ghemawat), a programming model that automatically partitions data, schedules tasks and handles machine failures — a blueprint that inspired open-source Hadoop and much of modern big-data tooling. Those innovations let Google scale services (Search, Maps, Books, Gmail) and amass unprecedented datasets and compute capacity, underpinning a business now worth roughly $3 trillion and generating $350B revenue in 2024, about 75% from advertising. The immediate news is legal: a U.S. federal court found Google a monopolist in general search and text advertising, ruling it violated the Sherman Act largely via exclusive distribution deals that lock Google as default search on many devices and browsers. Market shares cited were ~89% overall search (95% on mobile) and 88% of text ads. For the AI/ML community this matters — Google’s architecture and open publishing of core ideas accelerated industry-wide scaling, but the court’s ruling could force changes to default search distribution, ad revenue flows and ecosystem incentives. That may reshape access to data, default model APIs, cloud competition and business models that fund large-scale compute and training — with potential to open new competitive pathways or fragment datasets and infrastructure.