How ChatGPT could help Bing become relevant again

The much-maligned search engine Bing could soon have supremacy – thanks to AI.

Since Microsoft launched Bing in June 2009, the name has been a byword for embarrassment and ineptitude. The search engine was launched as Microsoft’s answer to Google’s dominance but quickly showed it was no match for the long-established leader of all things search.

Scrambling, Microsoft quickly signed a deal with Yahoo, one of Google’s longer-tenured competitors, to try and beef up the engine behind the search – but it too failed to make much of a dent in Google’s dominance.

Today, Microsoft sits a distant third in the race for search engine supremacy, with a market share of around 5%, far behind Google and the second-placed engine, China’s Baidu. But that could all change thanks to a series of recent announcements that could help Bing become more relevant again.

Bada-bing?

Following the release of ChatGPT by San Francisco artificial intelligence company OpenAI, the chatbot tool, which uses large language models (LLMs) to understand questions and converse with users in a semi-natural way, has become a major cultural force. It’s now used by 13 million users a day, according to investment bank UBS estimates.

And ChatGPT has attained a kind of cool cultural kudos that helps promote the idea that anything using its tools is automatically better and more intelligent than its competitors. The fact that Microsoft first invested $10 billion into ChatGPT creator OpenAI in late January, then announced in early February that it was integrating ChatGPT into its Bing search engine results, means that the Redmond, Washington-based company is stealing a march on its competitor.

There are plenty of questions still to be answered about whether Bing can ensure that the integration happens smoothly and in a way that benefits the company’s reputation for search. One of the things that meant Bing initially struggled to make an impact against Google when it was first released all those years ago was that the search results it served up were simply not very good. That was in part down to the comparative lack of power that the engine had in order to make things work.

The risk of false positives

Today, Microsoft faces a different challenge with Bing, but one that could have the same reputation-scuppering impact. Whereas in 2009, it was that the search results often were irrelevant, outdated, or simply not as good as Google’s, today they have to tackle the issue that blights ChatGPT and other generative AI tools like it: that it makes things up.

One of the major risks of using ChatGPT is the tool’s tendency to invent facts and present them with confidence when talking about something outside its training data. It will fabricate academic references if asked to write an essay and can get dates and details wrong – often key ones.

That’s a problem for a search engine, which deals in absolutes. People come to search engines because they’re seeking definitive answers to questions, and if they begin to think that they can’t trust what they’re being told through it, then they will soon learn how to use it again. That need to have absolute faith in the truth of any answers that a search engine throws up is why Google devotes so much time to trying to rid its results of spammy, SEO-bait content. If users feel unfulfilled when they use a search engine, they’ll seek out alternatives – and there are plenty of others out there.

It's for that reason that ChatGPT’s arrival on Bing could be a double-edged sword. On the one hand, it’s a potentially revolutionary addition to the search engine that could supercharge its performance and overhaul the way we seek out information online in a way that hasn’t been seen since Google introduced Page Rank way back when. But to pursue the high-reward strategy, you also have to take high risks. And ChatGPT’s tendency to hallucinate is the biggest risk of all.