The battle for AI-powered internet search has begun. ChatGPT’s success has ignited a hostile response from the current search powerhouse, Google, and Microsoft, which owns the Bing search engine. Richard Waters reports for the Financial Times:
The internet search wars are back.
The emergence of AI systems capable of generating direct textual answers to questions — most notably the ChatGPT chatbot created by San Francisco-based OpenAI — has opened the first new front in the battle for search dominance since Google fended off a concerted challenge from Microsoft’s Bing more than a decade ago.
Google and Microsoft are both close to announcing revamps of their search engines to include direct answers supplied by artificial intelligence. Meanwhile, several search start-ups have already embedded AI in their services, giving the first glimpse of how the technology behind ChatGPT might transform one of the biggest online markets.
The sudden spurt of experimentation is long overdue, said Greg Sterling, an analyst who has followed the search market since 1999. For younger users in particular, Google’s search results pages seem cluttered and strewn with advertising, he said. “People are ready for something that is simpler, seemingly more credible and doesn’t have tons of ads stuffed in it.”
On their own, systems such as ChatGPT, based on so-called large language models that can “understand” complex queries and generate text responses, do not represent a direct alternative to search. The information used to train ChatGPT is at least a year old and the answers it gives are limited to information already in its “memory”, rather than more targeted material pulled from the web in response to specific queries.
That has led to a race to develop a new hybrid of AI and traditional search. Known as retrieval augmented generation, the technique involves first applying search tools to identify the pages with the most relevant material, then using natural language processing to “read” them. The results are injected into a large language model such as OpenAI’s GPT-3, which then spits out a more precise answer.
Read more here.