In the race to build the infrastructure that powers artificial intelligence, Alphabet Inc.\u2019s Google has an enviable position: The company has a healthy cloud computing business, makes its own chips, and has struck deals to share them with companies like Anthropic PBC and Meta Platforms Inc. <\/p>\n
Google\u2019s success has made its computing resources so valuable, though, that its own AI researchers have to get in line. <\/p>\n
Last summer, Andrew Dai, then a researcher in Google\u2019s AI lab, discovered a blind spot in Gemini, the company\u2019s flagship AI model. While playing a board game, Dai took pictures of the board and asked Gemini a simple question: who\u2019s winning? To his surprise, Gemini was stumped, as were models from rivals. He became convinced of the need to build AI that could better understand what was happening in images.<\/p>\n
Dai discussed his idea with some of his colleagues, but he quickly concluded that he wouldn\u2019t be able to secure enough computing power to tackle the problem within Google, he said in an interview. He had to leave the company if he wanted to do it.<\/p>\n
Dai is among current and former employees who say Google\u2019s leadership in AI development has turned computing power into a precious resource, accessible mostly to people with high-priority projects, like improving Gemini.<\/p>\n
AI researchers sometimes feel like they are losing out on computing power to paying customers, the people said. Google\u2019s search and cloud computing units are also jockeying to use the company\u2019s chips, known as tensor-processing units, or TPUs. Within the AI lab Google DeepMind, access to computing power influences the projects that researchers pursue, the leaders they align themselves with and the pace at which they work.<\/p>\n
\u201cInside Google, every TPU has three suitors,\u201d said Oren Etzioni, a veteran AI researcher who is a professor emeritus at the University of Washington. \u201cIf you find yourself in the uncomfortable position where you have a pie-in-the-sky project and you are competing with a revenue-yielding customer, that\u2019s a tough position to be in.\u201d<\/p>\n
Google said in a statement that the company has a \u201crigorous, ongoing process that ensures our compute resources are allocated to the most important priorities, balancing today\u2019s customer and user needs along with our long-term investments to advance research and innovation.\u201d Alphabet Chief Executive Sundar Pichai has said that when deciding where to devote computing power, company leaders are focused on making sure that Google DeepMind has the resources that it needs to build cutting-edge AI models, \u201cbecause it\u2019s a foundation for everything we do.\u201d<\/p>\n
Alphabet said Google Cloud\u2019s backlog \u2014 the measure of contracted work that hasn\u2019t been recorded as revenue yet \u2014 nearly doubled from the prior quarter to over $460 billion. \u201cWe are compute constrained in the near term,\u201d Pichai said. \u201cWe are working through that moment and investing.\u201d Google will unveil its latest suite of product advancements at its annual developer conference in Mountain View on Tuesday.<\/p>\n
AI researchers once regarded Google as a place where they could have the freedom to pursue intellectual passions, almost like in academia, but with better pay and more resources. Researchers at the company have long angled for more computing power, but until relatively recently, the models were small enough that they didn\u2019t need as much to run a meaningful project, former employees said. But in 2022 the launch of OpenAI\u2019s popular chatbot ChatGPT prompted Google to invest in large language models, AI programs that can spin up a professional-sounding cover letter or term paper in seconds. Now Google is focusing on models that write computer code, which competitors have shown can be a hit product and generate revenue. <\/p>\n
Under the strategy followed by top AI labs, \u201cyou have to build the world\u2019s best coding model, because ultimately no one wants to be second to AGI,\u201d Dai said, referring to the widely held Silicon Valley ambition of building AI that can perform on par with humans. That makes the idea of pouring resources into other projects, especially experimental ones that may not generate revenue, harder for Google to justify.<\/p>\n
Dai left Google to found Elorian, an AI startup that recently exited stealth mode and specializes in visual reasoning, which Dai says is key to bringing AI to industries such as architecture, automotives and robotics. He is one of several former Google AI researchers who say they have had better access to computing power as startup founders. The researchers said that founding companies gives them the freedom to seek computing power from multiple sources \u2014 and they can use the chips they secure as they wish, without navigating Google\u2019s bureaucracy, or worrying access could disappear if company priorities shift.<\/p>\n
Former Google DeepMind researcher Ioannis Antonoglou said he had access to ample computing power while working on AlphaGo, an AI model designed to play the strategy game Go, which made waves by beating one of the world\u2019s best players. Later, he was part of the push to build Gemini, one of Google\u2019s most important strategic initiatives. But he felt that the company wasn\u2019t devoting enough computing power to post-training, a stage in which models are fine-tuned with data related to specific fields, such as legal documents or computer code. <\/p>\n
\u201cBoth myself and my cofounder, we believed in reinforcement learning as being the next frontier,\u201d said Antonoglou, who left with fellow DeepMind researcher Misha Laskin in 2024 to found ReflectionAI, a startup dedicated to building AI models in the open. \u201cIt wasn\u2019t clear that Google or DeepMind would take this path back then.\u201d <\/p>\n
When AI researchers are poised to defect, access to computing power is a lever that companies can pull. Former DeepMind researcher Anna Goldie said the company offered her more computing power to try to dissuade her from leaving to launch a startup. She ultimately departed anyway, founding a company called Ricursive Intelligence with fellow DeepMind researcher Azalia Mirhoseini that launched in late 2025.<\/p>\n
Goldie said she has been pleasantly surprised by how much computing power she has been able to find on the outside, from a range of sources. She declined to say how much computing power the company has obtained after raising $335 million, but she said it is on par with what she had been offered to stay at Google. <\/p>\n
\u201cI don\u2019t need to ask like 10 layers above me for permission,\u201d she said. \u201cI can just make a decision with my cofounder to do what\u2019s best for the company. I can listen to my employees and hear their ideas.\u201d<\/p>\n
At top AI labs, some researchers work on language models because it\u2019s the priority, even if their true interests lie elsewhere, said Tom McGrath, a researcher who left Google in 2023. <\/p>\n
\u201cThere\u2019s the carrot of compute and promo and generally being part of the glory of the big training run,\u201d said McGrath, who is chief scientist at Goodfire, a startup that aims to better understand the inner workings of AI models. \u201cThere\u2019s also the stick that you won\u2019t have any accelerators if you don\u2019t.\u201d <\/p>\n
It\u2019s a new way of life for some researchers at Google. To catch up in the AI race, Google in 2023 merged two AI labs: London-based DeepMind, which had a more top-down structure, and Google Brain, where researchers pursued passion projects with minimal supervision.<\/p>\n
Researchers at Brain each received credits to buy chips in an internal system where price fluctuated based on demand, similar to the stock market, Dai and Goldie said. Some researchers made the most of what they had by pooling resources and then using the credits of their teammates while they were on vacation or sleeping, Goldie added. \u201cThat was a powerful way that you could bond together and make something happen,\u201d Goldie said. <\/p>\n
Google still has a pool of computing power for individual researchers, but supply is constrained when the company is training large AI models, Dai said. This means researchers are effectively competing for slices of a smaller pie.<\/p>\n
Now, researchers who want more computing power often focus on short-term research questions that might yield something that could be incorporated into the next version of Gemini, Dai said. \u201cThen it makes leadership believe it makes more sense.\u201d<\/p>\n
Researchers can\u2019t always bank on receiving the computing power they are promised. In 2024, a large training run prompted Google to pause some research projects for about a quarter, Dai said. Some people abandoned their work as a result.<\/p>\n
Startups offer an \u201celement of control over your own destiny \u2014 being much clearer that if you pay for this much compute over the next year, you\u2019re going to get it,\u201d Dai said. \u201cNo one\u2019s going to take it away from you.\u201d<\/p>\n
To make the most of the computing power he has as Elorian ramps up, Dai said he has focused on hiring researchers who have experience with limited resources.<\/p>\n
\u201cThe game of AI has always been twofold,\u201d Antonoglou said. \u201cOne is, who has the most compute. And the second is, who can actually use it better.\u201d<\/p>\n
Love writes for Bloomberg.<\/p>\n","protected":false},"excerpt":{"rendered":"In the race to build the infrastructure that powers artificial intelligence, Alphabet Inc.\u2019s Google has an enviable position:…\n","protected":false},"author":2,"featured_media":42619,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[9],"tags":[9202,24,1201,6345,25984,532,25989,25985,25986,2408,25990,132,1429,25983,25987,25988,4835,1509],"class_list":["post-42618","post","type-post","status-publish","format-standard","has-post-thumbnail","category-google","tag-access","tag-ai","tag-ai-lab","tag-ai-startup","tag-andrew-dai","tag-company","tag-computing-resource","tag-enough-computing-power","tag-flagship-ai-model","tag-gemini","tag-goldie","tag-google","tag-google-ai","tag-google-researcher","tag-own-ai-researcher","tag-own-chip","tag-power","tag-revenue"],"_links":{"self":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts\/42618","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/comments?post=42618"}],"version-history":[{"count":0,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts\/42618\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/media\/42619"}],"wp:attachment":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/media?parent=42618"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/categories?post=42618"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/tags?post=42618"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}