It was Google's TGIF (Thank God It's Friday) all-hands this week, where employees get to put their most burning question to CEO Sundar Pichai and other company leaders. Top of mind for employees was AI, including the company's recent showcase of generative AI products at its I/O developer conference on May 10.
One employee said that after "weeks of flubs and embarrassing news stories," the response to I/O was "overwhelmingly positive." But staffers have concerns about the impacts of Google's aggressive push to build generative AI into all of its products.
"Many AI goals across the company focus on promoting AI for its own sake, rather than for some underlying benefit," read one of the top-voted questions. The employee asked how Google will "provide value with AI rather than chasing it for its own sake," according to a recording of the meeting reviewed by Insider.
Pichai responded at the meeting Thursday that when the company wrote its recent OKRs, it was during an "inflection point" around AI. OKRs are "Objectives and Key Results" — a system for setting goals used by Google and other tech companies.
These OKRs were written to "rally" teams to incorporate AI into Google's products, Pichai said.
"Clearly there's an implicit assumption and a conviction there that incorporating AI will make these products radically more helpful," Pichai told staff.
"But I do think the question is valid. I think one of the things the teams are all doing post-I/O is relooking," he said. "Normally we don't do this, but we are relooking at the OKRs and adapting it for the rest of the year, and I think you will see some of the deeper goals reflected, and we'll make those changes over the upcoming days and weeks."
'People don't want to just hear AI-generated content'
Another submitted question read: "LLMs are inundating the internet with AI-generated content, such as websites, books, images, music, and videos. This AI-generated content may eventually lead to a decline in internet quality, which may directly affect the quality of Google's search results. How are we addressing this challenge?"
Liz Reid, Google's vice president of search, responded by telling staff that they were focused on "surfacing really high-quality content."
"We think one of the things AI-generated content can do is enable people to make it easier for them to create content, to publish their ideas, to share their vision," she told staff.
She said the company is working on new techniques that focus on "understanding what is great content." She added that sometimes "great content" that Google surfaces will be AI-generated, but not always.
"Some of it has been great. It makes better weather and sports information. Some of it has been more spammy. And so that's what we're focusing on," said Reid.
"One thing we're seeing though, that I think we'll continue to reinforce is that people don't want to just hear AI-generated content, either from us or from the world at large," Reid said. "They often want to hear other people's perspectives, people they relate to, people's lived experiences."
"And so that's a big focus that we're doing, besides ensuring we don't show low quality, is figuring out how we elevate human voices in our products."
'Did Google decide there is no profit in shipping new non-AI products?'
A different question referred to a document that said Google didn't plan to buy new compute resources in 2024, aside from machine learning and cloud. "Did Google decide there is no profit in shipping new non-AI products?" the employee asked. "It will be hard to ship new products and make the efficiency improvements needed for organic growth on products like YouTube Shorts."
Ben Lutch, Google's engineering vice president, said that for work not in cloud or machine learning, Google is "starting now to focus engineering and product efforts on making sure we can grow and launch into our existing footprint" and wants to "make room for the highest-priority work." He noted that machine learning is "very power hungry" and the demand is growing rapidly.
"And so now is the time to clear out space and compress what we have so we can fit all the ML stuff as rapidly as possible into our data centers," Lutch said.
"This area is going to have intense work in the year ahead.," Pichai added. "The good news is there is a lot of exciting work we are doing which all needs compute, but constraint inspires creativity, and I think we will find the solutions we need to work through this moment as we've always done."
A Google spokesperson declined to comment.
