Suddenly, it seems like Google has been fooling around for the past decade.
The great world dominator of search engines has gone from being the leading technology company in artificial intelligence (AI) to seeming surpassed by Microsoft's new proposal in a few days.
The CEO of the latter company, Satya Nadella, presented last week a renewed Bing search engine, which will incorporate a chatbot developed by OpenAI, responsible for the famous ChatGPT.
Google counterprogrammed Microsoft by announcing Bard, its own version of a browser with smart chat, a day earlier.
But he was not able to show how it works, not even at a large event for the international press organized in Paris two days later, which was attended by EL PAÍS.
The only thing that could be seen there, in fact, took its toll on him: Bard's smart search recorded example gave incorrect information about the
Shares of Alphabet, Google's parent company, fell 8% that day.
The markets penalized that mistake, understanding that those from Mountain View are improvising a response to Microsoft's challenge.
Bard is an experimental conversational AI service, powered by LaMDA.
Built using our large language models and drawing on information from the web, it's a launchpad for curiosity and can help simplify complex topics → https://t.co/fSp531xKy3 pic.twitter.com/JecHXVmt8l
—Google (@Google) February 6, 2023
Why so much interest, all of a sudden, in AI?
Because ChatGPT has shown its potential to the general public.
Although the tool invents content, many thought that, with a few tweaks, it could revolutionize the search engine experience.
It is more pleasant to obtain information by talking to the machine than by typing keywords.
It is also interesting to be able to ask him to generate texts of a certain complexity, such as summaries, itineraries or essays.
Large language models (
) make this possible, although their reliability is still in question.
Some already believe that the hybridization of generative AI and conventional search engines may be the biggest innovation in consumer technology since Apple released its first iPhone.
Bing, which has always lived in the shadow of Google (3% and 90% of the world share of search engines, respectively), threatens for the first time the placid reign of the technology of colored letters.
The elephant in the room
But the frantic race to lead the development of increasingly intelligent search engines goes beyond riding a wave.
Controlling the world's most widely used search engine and web browser has allowed Alphabet and Meta to dominate the global advertising market for more than a decade, earning them an average revenue of $220 billion annually.
This rain of money has allowed him to buy strategic companies and launch the most diverse projects.
Among them, his autonomous car Waymo or Calico, the biotechnology company whose objective is to combat aging.
This bonanza may be coming to an end.
Last year was the first since 2014 in which Alphabet and Meta together accounted for less than 50% of the global advertising market, specifically 48.4%.
It is the fifth consecutive year that this figure has fallen since it reached its peak in 2017 (54.7%), and analysts expect that it will continue to decline.
The reasons: TikTok has been going strong, and is already the preferred search engine for many young people;
Amazon is growing, too, and Apple, since allowing app tracking to be blocked
has hurt Meta's business.
The great manna of advertising may be running out for Google and Facebook.
The second years ago that he decided what his response to that problem and the inability to attract a young audience was: the metaverse.
Google, for its part, has no plan B beyond AI.
It has been investing in this technology for decades.
That would explain his hasty reaction to Microsoft's bet.
a headlong race
Nadella has turned Microsoft around in less than a decade.
When the executive took control of the company in 2014, his income depended almost exclusively on Windows and the Office package.
He decided to bet big on cloud services and AI.
Azure, the cloud
, is already responsible for a quarter of the group's turnover.
Two years ago, Microsoft invested 1,000 million in OpenAI, to which this year, after verifying the tremendous success of ChatGPT among the general public, it has added another 10,000 to develop the conversational chatbot that will accompany its search engine.
What have you done while Alphabet?
Among other things, it has laid the foundations of the technology from which chatbots drink today, as the company's own executives have been careful to highlight lately.
Its Google Brain division and the British company DeepMind, which it acquired in 2014, are among the world's elite in the discipline.
As the CEO of the technology, Sundar Pichai, recalled last week, the Transformer research project and its founding article, presented in 2017, is the touchstone on which the scientific community has built the so-called advanced generative artificial intelligence.
Bard, Google's bet to make its search engine smart, is a pocket version of LaMDA, one of Google's most advanced language model projects.
Introduced two years ago, LaMDA grabbed international headlines last summer, when engineer Blake Lemoine, who was commissioned to review the ethical grounds of the robot's responses, said that, in his opinion, that artificial intelligence had gained consciousness.
DeepMind, for its part, plans to offer a beta version of its own model this year, which it has named Sparrow.
Denying the effect that the irruption of ChatGPT has had on the strategy of big technology companies is, at this point, unconvincing.
And yet, that is what Prabhakar Raghavan, vice president of Alphabet and one of the multinational's most powerful executives, did last week.
“We have been following our own roadmap in the development of artificial intelligence for years.
ChatGPT has not influenced us at all”, he assured on Wednesday in Paris in a meeting with various media, including EL PAÍS.
It is a fact, however, that Google has presented Bard, but without a release date.
Raghavan himself said that they did not have an approximate one: "What matters most to us is to achieve the quality that we want the service to have."
The technology industry is very given to move by trends.
Generative AI is clearly the
of the moment.
In addition to Microsoft and Google, Chinese tech giant Baidu also announced last week that it is working on its own version of an intelligent browser-chatbot hybrid.
Meta, for its part, canceled its Galactica project in November, a language model capable of producing scientific articles based on millions of previously analyzed documents, because it soon proved to be sexist and racist.
In order to prevail, search engines with chat must demonstrate that they provide reliable information.
It's not easy.
Examples of invented ChatGPT content have flooded social networks in recent months.
Bard inadvertently showed a mistake in his (the
telescope ) presentation at last week's event.
Bing, currently in the testing phase, also invents content if the screws are pushed.
Some of the world's leading experts warn against wanting to go too fast with this technology.
“Large language models should be used for writing aids, not much else,” said Yann LeCun, AI research lead at Meta and an eminence in the field.
Demis Hassabis, CEO of DeepMind (Google), also suggests that these tools require treading carefully: “It's okay to be cautious on this front,” he said.
That caution shines, for the moment, by its absence.
The limits of artificial intelligence
You can follow
EL PAÍS Tecnología
or sign up here to receive our
Subscribe to continue reading
Read without limits
I'm already a subscriber