Instead of trying to convert Google loyalists, Yahoo's AI search, Scout, aims to capture more activity from its own massive user base. The goal is to increase the search frequency of its 700M existing users, turning infrequent searches into a significant revenue stream.

Instead of relying solely on massive, expensive, general-purpose LLMs, the trend is toward creating smaller, focused models trained on specific business data. These "niche" models are more cost-effective to run, less likely to hallucinate, and far more effective at performing specific, defined tasks for the enterprise.

It's crucial to balance the hype around LLMs with data. While their usage is growing at an explosive 100% year-over-year rate, the total volume of LLM queries is still only about 1/15th the size of traditional Google Search. This highlights it as a rapidly emerging channel, but not yet a replacement for search.

Yahoo's new AI search engine, Scout, is built with a core value of sending traffic back to the open web via prominent links. This "blue link economy" approach is a strategic choice to differentiate it from rivals that summarize content, positioning Scout as an ally to publishers.

The cost to achieve a specific performance benchmark dropped from $60 per million tokens with GPT-3 in 2021 to just $0.06 with Llama 3.2-3b in 2024. This dramatic cost reduction makes sophisticated AI economically viable for a wider range of enterprise applications, shifting the focus to on-premise solutions.

Microsoft's research found that training smaller models on high-quality, synthetic, and carefully filtered data produces better results than training larger models on unfiltered web data. Data quality and curation, not just model size, are the new drivers of performance.

An emerging rule from enterprise deployments is to use small, fine-tuned models for well-defined, domain-specific tasks where they excel. Large models should be reserved for generic, open-ended applications with unknown query types where their broad knowledge base is necessary. This hybrid approach optimizes performance and cost.

Unlike chatbots that rely solely on their training data, Google's AI acts as a live researcher. For a single user query, the model executes a 'query fanout'—running multiple, targeted background searches to gather, synthesize, and cite fresh information from across the web in real-time.

Yahoo's new AI search engine, Scout, intentionally embeds direct, clickable links back to the original sources within its generated answers. This strategy aims to 'take care of the open web' by ensuring publishers receive traffic and credit, directly contrasting with other AI models criticized for scraping content without attribution.