Corporate Drive to Boost Productivity to Expand AI Use: Google Cloud CEO

Author: Tara Deschamps, The Canadian Press, July 2, 2024

Will Grannis, chief technology officer for Google’s cloud division, says that in the next year or two, many organizations will go from experimenting with AI to actually using the technology. Grannis poses at the Collision conference in Toronto, Wednesday, June 19, 2024. THE CANADIAN PRESS/Chris Young

Google’s cloud chief technology officer says that in the next year or two, many organizations will move from experimenting with AI to actually using it. Will Grannis says that as companies emerge from trial mode, more will turn to AI-based platforms and tools for everything from financial services to healthcare. He believes that the shift will be driven by the world’s growing tech savvy and the ongoing drive to improve productivity and efficiency, particularly in the workforce. “Public sector, private sector, commercial—it doesn’t matter because we all want to run our businesses more efficiently,” Grannis said in an interview during a recent trip to Toronto for the Collision technology conference. “And what we’re finding today is that there are a lot of things that people do manually that don’t add much value.” Canada’s productivity rate—the amount a country produces for each hour worked—has fallen in recent years to a level that’s now 30 percent lower than the U.S., according to a June 20 report from the Royal Bank of Canada. Senior Deputy Governor of the Bank of Canada Carolyn Rogers went so far as to call the trend a national emergency in a speech in March. Data cited by RBC suggests that AI could reverse the trend, potentially saving every worker in the country between 100 and 125 hours a year and increasing productivity by eight per cent by 2030. Customer service and software development are areas that are particularly ripe for AI, Grannis said. He’s heard of some Canadian support centres that are fielding up to 70,000 calls a day, mostly handled by humans. The introduction of AI to those centres means agents answering calls could be given information about a customer’s history or the services they’re using. They could also get help translating between languages. In software development, many companies employ engineers to build apps and other products. “The first thing most software engineers do is try to find something that looks like (what their company wants), and then copy and paste it, and then modify it,” Grannis said. With AI, they could ask a model to write code that describes exactly what they want to achieve. The AI ​​would be able to perform the task using any programming language the programmer wants, and engineers could ask another form of technology to even critique the work of the first. A third could fix any problems it finds. “You use AI in this kind of workflow management and you get the benefits of that,” Grannis said. “Now the software engineer can use their creativity, they can use their domain knowledge, and they can get those software drafts out 10, 100 times faster.” A June report from Microsoft found that programmers who used generative AI tools could complete tasks 56 percent faster than those who didn’t, and those who used the technology for writing could reduce their time spent on their jobs by 37 percent. But many worry that the growing reliance on AI for such tasks will contribute to unemployment. A 2020 Statistics Canada study found that 10.6 percent of Canadian workers were at high risk of having their jobs transformed by automation, while another 29.1 percent were at moderate risk. That transformation could include anything from losing their job to a complete overhaul of their responsibilities. Grannis said that to navigate such a transformation and reskill for the changing job market, workers will need to get comfortable with AI as soon as possible. “It can’t just be using an app or taking an online course,” he said. “It requires a certain amount of curiosity.” Younger people, he added, already have that curiosity. They’re using the technology to write drafts of articles or code and find software vulnerabilities that they have to deal with during cybersecurity internships. “So in many ways, it makes other cohorts and other demographics more comfortable,” Grannis said. Getting people comfortable also means helping them understand the limitations of the technology. There are still many things that AI can’t do, and even what it can do isn’t always perfect. AI has been known to hallucinate—to provide false or misleading information based on data that it believes to be true but isn’t. For example, Google’s chatbot Bard claimed last year that NASA’s James Webb Space Telescope had taken the first images of a planet outside our solar system. Those images were actually taken by the European Southern Observatory’s Very Large Telescope in 2004. Asked whether AI will ever get rid of all its problems and reach a pristine state, Grannis said: “Well, AI is created by humans.” “Humans are not pristine,” he said. “So I assume there will always be things that need to be worked on to improve AI.” This report by The Canadian Press was first published July 2, 2024. 32
-31