Editorial MrIncredible/iStock via Getty Images
Google (NASDAQ: GOOG) (NASDAQ: GOOGL) and Microsoft (NASDAQ: MSFT) have taken over the majority of investor opinion in the battle for generative AI possibilities, as both tech titans make changes to search and use it chatbots into their respective products.
While Google ( GOOG ) ( GOOGL ) isn’t gunning for user numbers or market share, a deeper-than-expected integration of artificial intelligence could end up costing the company significantly in the short term, according to Morgan Stanley analyst Brian Nowak said.
Nowak, who has an overweight rating and a $135 price target on Google ( GOOG ) ( GOOGL ) stock, noted that if 10% of search moves to language models, it could add as much as $1.2 billion in operating costs for Sundar A company run by Pichai.
Microsoft ( MSFT ) revealed this week that it sees roughly 10 billion search queries per day, which Nowak said would suggest that Google ( GOOG ) ( GOOGL ) sees about 3.3T per day, giving Microsoft roughly 3% market share, according to StatCounter.
While the 3.3T queries are far lower than estimated, this means that Google ( GOOG ) ( GOOGL ) is spending roughly one-fifth of a cent per search query. Nowak estimates that natural language queries could cost five times more than traditional queries, given the high level of computational intensity for natural language models.
Nowak went so far as to say that if Google ( GOOG ) ( GOOGL ) sees half of its queries using natural language integration, it could increase the company’s costs by as much as $6 billion and reduce earnings before interest and taxes by up to 6%.
As such, Nowak suggests there are four factors to look out for, including the speed at which Google is rolling out its consumer tools and the quality of those tools.
Nowak said that so far, monthly trackers comparing ChatGPT, Bing and Google show “no material signs” of Google weakness. According to Nowak, this situation “speaks to the likely high incrementality of these [natural language] search queries”, and that the quality, accuracy and reliability of the results will be important to drive “consistent” appeal.
That could be a problem in the near future for Google ( GOOG ) ( GOOGL ), as Bard gave an incorrect result in the company’s recent announcement, when it said that the James Webb telescope had taken the first images of planets outside the solar system, known as exoplanets.
The first exoplanet was discovered in 1992, and more than 5,000 have been discovered to date. The James Webb Telescope was launched on December 25, 2021.
Despite the early misstep, Nowak said Google’s ( GOOG ) ( GOOGL ) scale and faster feedback loops will likely be a “positive” for it.
Other key factors to look at include developer buy-in, given the initial state of generative AI. It’s likely that more AI tools will be used in next-gen products, but it’s still early days for them, and given the buzz around ChatGPT, it’s possible that there will be overlap when it comes to the developer community.
Nowak said there is also the possibility that artificial intelligence can result in more purchasing transactions in fewer steps. Such a change could see transactions run on Google ( GOOG ) ( GOOGL ) , resulting in “incremental revenue” for the tech giant while excluding third parties such as small e-commerce companies, local businesses like Yelp (YELP) and online travel agencies.
Earlier this week, Google ( GOOG ) ( GOOGL ) showed off a new multi-search feature in Google Lens that lets people search for items using text and images and is useful for shopping. The feature, previously available in the US and India, is now available to mobile users worldwide, Google said.