Google warns of a ‘paradigm shift’ in AI that companies can’t handle alone

Google CEO Sundar Pichai said on Wednesday that the rapid development of artificial intelligence (AI) represents a “paradigm shift” that needs to be approached responsibly and in collaboration between businesses.

“We’re doing it responsibly to get things right,” Sundar Pichai said during a kickoff presentation at Google’s annual I/O event in Mountain View. “No company can do this alone.”

The CEO called for collaboration between companies to cope with this rapid evolution, at an event that focused almost entirely on artificial intelligence and its applications in the short term.

“I’ve been thinking about the huge technological shifts that we’ve all been part of, and the shift in AI is as big as you can imagine,” said Sundar Pichai. “That’s why it’s so important to make AI useful for everyone.”

In a two-hour presentation, Google’s senior vice president of technology and society, James Manica, explained that there are many risks associated with the potential of artificial intelligence, which has seen an exponential jump in applications available to consumers over the past six months. .

“We have to acknowledge that this is a new technology that is still under development and there is still a lot to be done,” Manica said.

Sundar Pichai described Google’s approach as “brave and responsible”, while Manika emphasized that there is a natural tension between these positions.

“The only way to be truly courageous in the long run is to be accountable from the start,” he said.

Google recently introduced its AI-based chat bot Bard, which is seen as a response to the popularity of OpenAI/Microsoft’s ChatGPT.

From large language models (LLMs, large language models), systems such as ChatGPT and Bard can generate content based on user requests in multiple languages. This is what is commonly referred to as generative AI.

Francisco Geronimo, deputy vice president of IDC’s EMEA devices division, told Lusa that caution makes sense.

“This is true. Because this is a technology that no one knows about until now, what positive and negative potential it has, ”he explained. “Sundar, like other executives, knows that this technology can be very dangerous.”

According to what he presented at I/O, Google is expanding the capabilities of generative AI across several of its products and services for both companies and consumers, but calls for standardization of security measures to mitigate risks.

“Despite the fact that it is interesting, AI can also exacerbate unfair social prejudices,” Manika cites the example.

One of Google’s biggest problems is the rise of misinformation, which is why the company is developing tools to verify the authenticity of information.

“We will make sure that generative AI images contain metadata in the source file,” Manica assured. Creators and editors will be able to embed metadata and watermarks into files, which users can then use to understand what is genuine and what has been synthetically created.

Another example is technology that automatically dubs speech in a video into another language and edits the image so that the lips move according to the new language.

“This is a huge step forward. But he has internal tensions,” Manika said. This technology can be used by third parties with bad intentions to create “deepfakes,” he explained, referring to videos that artificially reproduce a person and their voice to make them say something they didn’t actually say.

Google has said the technology will only be available to authorized partners to prevent it from being used for nefarious purposes.

“The responsible creation of AI must be a collaborative effort involving researchers, scientists, industry experts, governments and people in everyday life,” Manica said. “We have a lot to do together.”

Francisco Jeronimo liked what he saw, although he noted that none of the new products were revolutionary yet.

“At this point, Microsoft continues to outpace Google, trying to dominate this space,” he said.

Sundar Pichai said that the company is guided by the principles of responsible AI, and at various times pointed out that Google is a pioneer in the application of artificial intelligence technologies.

The mention was due to the fact that Microsoft, which has invested in OpenAI, is now credited with being the most advanced approach in this age of AI by integrating the technologies that ChatGPT is based on into their products.

Google I/O is taking place in Mountain View, at the company’s headquarters, and also serves as a showcase for several new products, including a tablet and a foldable smartphone.