New AI threatens Google’s dominance in search

THE NEW YORK TIMES – Over the past three decades, products like Netscape’s browser, the Google and the Apple iPhonesreally aimed at technology sector upside down and made those in front of them look like heavy dinosaurs.

A few weeks ago he called an experimental chatbot (software that can “talk” to humans). Chat GPT he set out to be the next big game changer in the industry. It can provide information in simple, clear sentences rather than just a list of internet links. Can explain concepts in ways that people can easily understand. It can even generate ideas from scratch, including business strategies, Christmas gift suggestions, blog topics, and holiday plans.

While ChatGPT still has a lot of room for improvement, its release has led Google management to declare it a “code red”. For Google, it was like setting off a fire alarm. Some worry that the company is approaching a moment that the biggest companies in Silicon Valley fear: the arrival of a huge technological shift that could turn the business upside down.

For more than 20 years, the Google search engine has been the world’s leading Internet portal. But with a new generation of chatbot technology poised to reinvent or even replace traditional search engines, Google could face the first serious threat to its core search business. A Google executive described the efforts as critical to Google’s future.

ChatGPT was launched by an aggressive research lab called Open AIand Google is among many other companies, labs and researchers who have helped develop this technology. But experts believe the tech giant may find it difficult to compete with the smaller, newer companies building these chatbots due to the many ways this technology can harm their business.

Google has spent several years working on chatbots and, like other big tech companies, has been aggressively pursuing AI technology. Google has already built a chatbot that can compete with ChatGPT. In fact, the core OpenAI chatbot technology was developed by researchers at Google.

Called LaMDA, or Language Model for Dialogue Applications, the Google chatbot received a lot of attention over the summer when a Google engineer, Blake Lemoine, claimed to be sentient. This was not true, but the technology has shown how much chatbot technology has improved in recent months.

Blake Lemoine was fired by Google after claiming the company’s AI was conscious

However, Google may be reluctant to roll out this new technology as a replacement for online search because it’s ill-suited to serving digital ads, which accounted for more than 80% of the company’s revenue last year.

“No company is invincible; everyone is vulnerable,” said Margaret O’Mara, a professor at the University of Washington who specializes in the history of Silicon Valley. “For companies that have had tremendous success doing one thing that defines the market, it’s hard to have a second act with something totally different.”

As these new chatbots learn their skills by analyzing massive amounts of data posted on the internet, they manage to blend fiction and reality. They provide information that can be biased against women and people of color. They can generate toxic language, including hate speech.

All of this could turn people against Google and damage the corporate brand it took decades to build. As OpenAI has shown, new companies may be more willing to risk claims in exchange for growth.

Continue after commercial

Even as Google perfects chatbots, it faces another problem: Does this technology cannibalize the company’s lucrative search ads? If a chatbot answers questions in short sentences, there are fewer reasons for people to click on advertising links.

“Google has a business model problem,” said Amr Awadallah, who worked for Yahoo and Google and now runs Vectara, a startup developing similar technology. “If Google gives you the perfect answer to every question, you won’t click on any ads.”

Sundar PichaiCEO of Google, he was involved in a series of meetings to define Google’s AI strategy and was instrumental in the work of several groups within the company to respond to the threat posed by ChatGPT, according to a memo and audio obtained from the The New York Times. Employees have also been tasked with creating AI products that can create artwork and other images, such as OpenAI’s DALL-E technology, which is used by more than 3 million people.

Sundar Pichai, CEO of Google, at the company's headquarters in Mountain View, California, USA
Sundar Pichai, CEO of Google, at the company’s headquarters in Mountain View, California, USA Photography: Max Whittaker/The New York Times

Between now and a major conference Google is expected to hold in May, Google’s research, trust and security teams and other departments have been redeployed to help develop and launch new AI prototypes and products.

As technology advances, industry experts believe, Google must decide whether to overhaul its search engine and turn a full-fledged chatbot into its flagship service.

Continue after commercial

Google is reluctant to share its technology widely because, like ChatGPT and similar systems, it can generate false, toxic, and distorted information. LaMDA is only available to a limited number of people through an experimental app, AI Test Kitchen.

Google sees this as a struggle to implement its advanced artificial intelligence without harming users or society, according to a memo seen by the Times. At a recent meeting, a manager acknowledged that smaller companies have fewer concerns about implementing these tools, but said Google must join the fray or the industry can get along without it, according to an audio recording from the company. meeting obtained. from the Times.

Other companies have a similar problem. Five years ago, Microsoft released a chatbot called Tay that spewed racist, xenophobic and obscene language and was forced to immediately remove it from the internet, never to return. In recent weeks the Objective removed a new chatbot for the same reasons.

Executives said in the taped meeting that Google intended to roll out the technology powering its chatbot as a cloud computing service for outside companies and could incorporate the technology into simple customer service tasks. It will maintain its trust and safety standards for official products, but will also release prototypes that don’t meet those standards.

It can limit these prototypes to 500,000 users and warn them that the technology can produce false or offensive claims. Since its launch on the last day of November, ChatGPT – which can produce similar toxic material – has been used by more than 1 million people.

“A great example of a conversation system people can interact with within a few rounds that sounds mind blowing? It’s a good step, but it’s not what will really transform society,” said Zoubin Ghahramani, who oversees Google’s Brain AI lab, in an interview with the Times last month ahead of ChatGPT’s launch. “It’s not something people can reliably use on a daily basis.”

Continue after commercial

Google is already working on improving its search engine using the same technology that powers chatbots like LaMDA and ChatGPT. Technology – a “big language model” – is not just a way for machines to carry on a conversation.

Today, this technology helps Google’s search engine highlight results that aim to directly answer a question you’ve asked. In the past, if you typed, “Do beauticians stand up a lot at work?” on google, it didn’t understand what you were asking. Now, Google answers that correctly with a short description describing the physical demands of life in the skincare industry.

Sridhar Ramaswamy, head of Google's advertising business, including search ads, is one of the leaders of the new Neeva search technology.
Sridhar Ramaswamy, head of Google’s advertising business, including search ads, is one of the leaders of the new Neeva search technology. Photography: Jason Henry/The New York Times

Many experts believe Google will continue to take this approach, gradually improving its search engine rather than overhauling it. “Google search is pretty conservative,” said Margaret Mitchell, who was an AI researcher at Microsoft and at Google, where he helped start its ethical AI team, and is now at the Hugging Face research lab. “Try not to disrupt a system that works.”

Other companies, including Vectara and a search engine called Neeva, are working to improve search technology in similar ways. But as OpenAI and other companies improve their chatbots, working to address issues like toxicity and bias, it could become a viable replacement for today’s search engines. Whoever arrives first can be the winner.

“Last year I was disheartened that it was so difficult to remove Google’s iron grip,” said Sridhar Ramaswamy, who previously oversaw Google’s advertising, including search ads, and now runs Neeva. “But technological moments like this create an opportunity for more competition.” / TRANSLATION LÍVIA BUELONI GONÇALVES

#threatens #Googles #dominance #search

Add Comment