Saturday, December 9, 2023

Did Google hide its advances in generative AI to maintain its search engine monopoly?

For several weeks now, Google faces the most important antitrust trial in its history. The United States Department of Justice is leading this process, where it seeks to demonstrate that those in Mountain View engaged in anti-competitive practices to have practically absolute control of web searches.

And the case is notable enough to shed light on details that under normal conditions ordinary people would not know. The most notable, without a doubt, is how much Google pays Apple to remain the default search engine in Safari. But he’s not the only one.

In fact, this week the Department of Justice leveled one of the strongest accusations against Californians, and it went practically unnoticed. According to the authorities, Google hid its advances in generative artificial intelligence so as not to harm the monopoly it exercises through its search engine.

That Google has been dedicating time and resources to the development of artificial intelligence models for years is nothing new. The same applies to the fact that the company chose to gradually incorporate this technology both in its search engine and in its other products. One of the clearest examples is that of LaMDA. This model was presented in 2021, but was quickly forgotten. And it was only talked about again when an engineer caused a stir by claiming that he had become conscious.

The US believes that Google tried to hinder its advances in generative AI

It is also evident that this year Sundar Pichai’s team has redoubled its efforts to implement and promote AI-based tools. Both at the I/O event in mid-May and at the recent presentation of the Pixel 8/8 Pro and the Pixel Watch 2, the words “artificial intelligence” They were repeated ad nauseam.

It is on this point where the Department of Justice wants to play its strongest cards. The organization maintains that, if it were not for the emergence of OpenAI with ChatGPT and Microsoft’s push to take advantage of that technology, many of the AI-related advances that Google has presented this year—call them Bard, PaLM 2, or whatever— they would not have seen the light yet.

The launch of ChatGPT at the end of 2022 caused a real sensation and put generative artificial intelligence center stage. Microsoft jumped at the opportunity to give OpenAI several billion dollars to integrate the technology into its products, starting with the Bing search engine. And Google had no choice but go out and answer him from the hand of Bard and various other advertisements.

According to the DOJ, Google had the economic means and the necessary technical advances to launch its tools based on generative AI much sooner.. However, he considered that it was not until the competition first hit the market that those from Mountain View decided to make public their efforts in the matter.

The authorities believe that, if it had depended on Google, all the advances seen in the last 10 months they would not have come to light. Or at least not with such speed. And this is where an important part of the argument behind the antitrust lawsuit against the Californians lies.

From Mountain View they defend themselves

Photo by Pawel Czerwinski on Unsplash

The tactic of the United States Department of Justice is not new. As well indicated Fortune, already used a very similar one in the lawsuit that in 1982 ended the monopoly that AT&T exercised over the American telephone system. Under the same premise, if it is now verified that was weighted on purpose developing technologies that would be beneficial to the public, the story could quickly get complicated for Google.

Of course, those from Mountain View are not leaving this point to chance. Days ago he testified Prabhakar Raghavan, the leader of the team in charge of Google’s search engine. The executive asserted that they did not hinder the development of generative AI to maintain the monopoly of their search engine. On the contrary, they chose to take slow, but firm steps in their approach.

“We felt it was not yet responsible to introduce that technology to users due to concerns about its feasibility and toxicity. We kept it behind the scenes, but gradually developed it,” she explained.

It is evident that this discussion is part of a much broader debate and that There will be no immediate resolution.. A verdict in the antitrust trial against Google is not expected to come until well into 2024.

Also in Hipertextual:


Please enter your comment!
Please enter your name here

Latest article