A week of clashes between Microsoft and Google in this exciting new phase of the search engine wars has made Redmonds aware of the David vs. Goliath battle they face.
Bing, Microsoft’s search engine, has been revamped with artificial intelligence superpowers. Google’s namesake search engine is going to get the same treatment. Both aim to deliver better, more conversational and engaging search results, rather than endless lists of links.
The launch of the “new Bing” last week was greeted with great anticipation. In a blog post on Wednesday, Microsoft claimed that Bing had received user approval for 71% of responses in a large-scale public test.
These positive results reinforce Microsoft’s conviction in its multi-billion dollar bet on OpenAI, creator of ChatGPT, which develops the AI that powers the new Bing.
That said, Microsoft may have overestimated the impact the “new Bing” technology will have. For one, its launch hasn’t gone entirely to plan, and even Microsoft has admitted that Bing has misbehaved in certain circumstances.
And besides, Bing faces a powerful enemy: Google.
There’s no doubt that Microsoft execs got a bit of a kick out of it when Google introduced Bard, its ChatGPT rival, last week, and the chatbot got its launch presentation wrong. Google’s market capitalization fell by $100 billion after the event.
Considering that Tay, a former Microsoft chatbot, went wild and gained notoriety for all the wrong reasons, the company is well aware that the new Bing runs just as much risk if it doesn’t measure up.
But independent technical engineer Dmitri Brereton pointed to several mistakes made by Bing in its demo last week, suggesting that “no one noticed” at the time because “everyone jumped on the Bing euphoria bandwagon.”
When asked about a vacuum cleaner, Bing produced a list of pros and cons that Brereton said was based on “completely fabricated information.” And when Bing confidently offered citations, they didn’t match the information, according to Brereton.
A more egregious example shared by Brereton was a financial statement summary for clothing maker Gap, which included an inaccurate reading of its gross margin and a “completely fabricated” figure for its operating margin. oops.
For Victor Botev, co-founder and CTO of European startup Iris.ai, Microsoft will have to do much more than offer plausible answers if it wants to take on Google in the search engine wars.
Botev believes that the new Bing and the new Google to come are failing to answer specific queries. The more “technical, domain-specific, and scientific” the query, the more likely “flaws in the training data and back-end model” will be exposed by these new search engines.
“For more specialized fields, whether in business, scientific research, or beyond, the stakes are too high for anything other than absolute confidence in the answer,” Botev added in his reflection.
David looks at Goliath
Currently, Google’s share of the search market is almost 93%, according to data company StatCounter. Bing’s market share is around 3%. It’s a chasm that doesn’t look like it will narrow much if Bing’s teething problems persist.
Microsoft corporate vice president Philippe Ockenden told analysts last week that Bing’s advertising business could earn $2 billion in revenue “for every point of share gain in the search advertising market.”
Gross ad revenue for Google — the growth engine of search — last year was $224 billion, compared with about $18 billion for Microsoft, according to a study by Jefferies analysts.
It is quite possible that the new Bing will gain more fans once it is more widely deployed. And that would mean more ad money for Microsoft.
That being said, Microsoft will be well aware of the reputational damage that could be caused by unleashing an indomitable AI, which is what Bing has proven to be in its current form. If that’s the case, those extra ad dollars won’t mean a thing.
So right now, the new Bing looks like an exciting technology that promised too much before falling short. Which means, for now, it’s not Google’s giant killer.