Artificial intelligence is supposed to be something of Google. The company has a reputation for making long-term bets on all sorts of far-fetched technologies, and much of the research that underpins the current wave of AI-powered chatbots has taken place in the company’s labs. However, a startup called OpenAI soon emerged as a leader in so-called synthetic AI software—which can create its own text, images, or videos—by launching ChatGPT. in November. Its surprise success sent Google parent Alphabet a sprint to catch up with a key technology area that CEO Sundar Pichai has said will be “deeper than fire or wind.” electricity”.
ChatGPTwhich some see as the ultimate challenger to Google’s traditional search engine, seems doubly threatening due to OpenAI’s close ties to Microsoft. feeling that Google According to current and former employees as well as others close to the company, many of whom requested anonymity may be lagging behind in an area where it has been seen as a key strength. led to no small amount of anxiety in Mountain View, California. because they are not allowed to speak in public. As one current employee put it: “There’s an unhealthy mix of unusually high expectations and great insecurity for any AI-related initiative.”
Effort has been Pichai relive his days as a product manager, when he had to weigh up the details of a product’s features first-hand, a task often far below his salary, according to one former employee. Google co-founders Larry Page and Sergey Brin have also been more involved with the company than in previous years, with Brin even submitting code changes to Bard, ChatGPT-esque chatbot from Google. Senior management has declared a “red code” that comes with the directive that all of their most important products — those with more than a billion users — must incorporate general AI within a few months, according to the company. someone with knowledge of the matter. In an early example, the company announced in March that creators on YouTube Video platforms will soon be able to use this technology for virtual costume swaps.
Several Google alumni were reminded of the last time the company took on an internal task of infusing every major product with a new idea: an effort that began in 2011 to promote the ill-fated social network. Google+. It’s not a perfect comparison—Google has never been seen as a leader in social media, while its expertise in artificial intelligence is indisputable. However, there is still a similar feeling. Bonuses for employees have been associated with the success of Google+. Current and former employees say at least some Google employee ratings and reviews will likely be affected by the ability to integrate general AI into their work. Code red has resulted in dozens of generalized AI integrations already planned. “We’re throwing spaghetti against the wall,” said one Googler. “But it’s not even close to what it takes to transform the company and compete.”
In the end, the mobilization around Google+ failed. The social network struggled to find traction with users, and Google finally announced in 2018 that it would be shutting down its consumer product. A former Google executive sees failure as a cautionary tale. This person said: “Larry’s mission is that every product must have a social component. “It ended pretty badly.”
A Google spokesperson pushed back against comparisons between code red and Google+ campaigns. While Google+’s mission spans all products, the current AI push largely consists of Google employees being encouraged to test the company’s AI tools internally, the spokesperson said. says: a common practice in tech is nicknamed “dogfooding”. The spokesperson said most Google employees don’t spend more time on AI, just those working on related projects.
Google is not alone in believing that AI is now everything. Silicon Valley has entered a full-blown hype cycle, with venture capitalists and entrepreneurs suddenly declaring themselves AI visionaries, away from recent fixes like the Blocks and companies saw their stock prices skyrocket after announcing AI integrations. In recent weeks, Meta . Platform CEO Mark Zuckerberg has focused on AI rather than the metaverse — a technology he recently claimed was so fundamental to the company that it required a name change, according to two people familiar with the matter.
The new marching orders are welcome news to some at Google who are well aware of the history of diving into speculative research only to stumble upon it commercializing. Members of several teams that have worked on innovative AI projects hope that they will now be able to “ship more and have more products, rather than just some kind of research.” according to one of the people with knowledge of the matter. .
In the long run, it may not matter much openAI has sucked all the air out of the public conversation for several months, based on the amount of work Google has completed. Pichai began calling Google an “AI-first” company in 2016. Google has been using machine learning to drive its advertising business for years, while also embedding AI in key consumer products such as Gmail And Google Photos, where it uses technology to help users compose emails and organize images. In a recent analysis, research firm Zeta Alpha examined the 100 most cited AI research papers from 2020 to 2022 and found that Google dominates the field. “Ultimately, it shows that Google is a sleeping giant, now in the background, playing catch-up. I think that’s not quite the case,” said Amin Ahmad, a former AI researcher at Google who co-founded Vectara, a startup that provides conversational search tools to businesses. “I think Google has been really good at applying this technology to some of their core products years ahead of the rest of the industry.”
Google has also struggled with the tension between its commercial priorities and the need to handle emerging technology responsibly. There is a well-documented trend in automated tools that reflect biases that exist in the datasets they have been trained on, as well as concerns about the impact of experimental tools. to the public before they are ready. Innovative AI in particular comes with risks that make it impossible for Google to rush to market. For example, in search, a chatbot can give a single answer that appears to come directly from the company that created it, similar to how ChatGPT appears to be the voice of OpenAI. Basically, this is a riskier proposition than providing a list of links to other websites.
Google’s red code appears to have scrambled risk-reward calculations in a way that has alarmed some experts in the field. Emily Bender, a professor of computer linguistics at the University of Washington, said Google and other companies catching on to the broad AI trend may not be able to steer their AI products away from “false examples.” the most serious, let alone the prevailing trend but slightly more sophisticated cases.” The spokesperson said Google’s efforts are governed by its AI principles, a set of guidelines published in 2018 for developing the technology responsibly, adding that the company is still taking a cautious approach.
Other outfits have shown they’re willing to push, whether Google does or not. One of the most important contributions Google researchers have made to the field is a landmark paper titled “Attention Is All You Need,” in which the authors introduced transformers: systems that help AI models focus on the most important pieces of information in the data they are analyzing. Transformers are now the main building block for large language models, the technology that powers the current group of chatbots—the “T” in ChatGPT stands for “transformer.” Five years after the article was published, all but one of the authors left Google, for some reason wanting to escape the rigors of a large, slow-growing company.
They are among dozens of AI researchers who have turned to OpenAI as well as a host of smaller startups, including Character.AI, Anthropic, and Adept. Several startups founded by Google alumni—including Neeva, Perplexity AI, Tonita, and Vectara—are looking to reimagine search using massive linguistic models. Sara Hooker, a Google Brain alum, said the fact that only a few key places have the knowledge and the ability to build them makes the competition for that talent “much more intense than across sectors.” but the method of model training is not as specialized as”. currently working at AI startup Cohere.
It is not uncommon for people or organizations to make significant contributions to the development of one disruptive technology or another, only to witness others realize the incredible financial benefits that come with it. Without them. Keval Desai, a former Googler who is now the chief executive officer of venture capital firm Shakti, cites the example of Xerox Parc, the research lab that laid the groundwork for much of the personal computer era. , only to witness Apple Inc. and Microsoft came along and built their trillion dollar empire on their backs. “Google wanted to make sure it wasn’t the Xerox Parc of its day,” Desai said. “All the innovation happened there, but there was no execution.”
© 2023 Bloomberg LP
