IMAGE: Google
The race between Microsoft and Google to be the first to include a machine learning conversational assistant as part of its search engines is hotting up: in a post on its corporate blog signed by CEO Sundar Pichai, Google has announced the launch of Bard, for the moment in test mode restricted to a small group of users, an assistant which is similar to ChatGPT, and based on its Language Model for Dialogue Applications (LaMDA). This is clearly a rushed response to the success of ChatGPT.
In response, Microsoft held a surprise in-person event on Tuesday at 1900 CET to announce the inclusion of ChatGPT into Bing, which will most likely look like what we’ve seen from screenshots leaked by users who stumbled upon it on Friday.
The race to put the product in the hands of users has been accompanied by all kinds of subterfuge: leaked screenshots, announcements, events, and all kinds of PR in a battle of egos to capitalize on the competitive advantage of being the first. Microsoft wants to score the point of having been visionary and for working with OpenAI since its inception to be able to beat Google in its field, while Google does not want to remembered as the company that knew the most about machine learning, but allowed a minor search engine to snatch the honor of being the first to integrate it in an efficient and appealing way for users.
As ever, the proof of the pudding will be in the eating. In Microsoft’s favor, many users have already had time to familiarize themselves with ChatGPT, but with the problem of launching a product integrated into a search engine that is used by a very small percentage of people In Google’s favor, its huge market market share and global brand awareness of its search engine.
What will be the consequences of such a significant change in the way we see Google results? What happens if, overnight, a significant number of people are satisfied with Google’s text response and stop clicking on the algorithm’s natural results? For millions of brands that have been working for years on their position in those results listings, this could cause chaos: they may head the results of searches, but no one cares anymore, because the ones that count are the ones that Bard takes and converts into a text with which it responds to the user. Type 1, navigational, searches might still work; but type 2, transactional, searches might become less automatic and more unpredictable depending on what Bard answers (imagine, for example, questions like “which digital camera should I buy”). But type 3, informational searches, would change, and the consequences will be hard to foresee. I can think of quite a few SEO professionals who will be scratching their heads about this.
On the screen that Google has advanced, the integration of its conversational algorithm seems to be immediately below the search box, in the same place where it has positioned other products such as comparators, making it hard to know whether the highlights immediately below correspond to the sources used to elaborate the answer or are simply additional search engine results in “read more” mode.
Regardless of who gets their product out there first, this is one arms race that looks set to continue for some time.
(En español, aquí)