How much money do the world's tech giants spend on artificial intelligence technology?
With the potential to change the world, technology giants such as Google, OpenAI, Meta... are constantly pouring money into research and development of artificial intelligence (AI) technology, turning this into an uncompromising race.
The AI race is extremely fierce as technology giants are constantly pouring money into this field. Accordingly, Microsoft has invested 13 billion USD in artificial intelligence startup OpenAI. Other technology giants such as Amazon and Google have also poured 4 billion USD and 2 billion USD respectively into artificial intelligence startup Anthropic, a direct competitor of OpenAI.
Meanwhile, OpenAI itself has been actively investing in promising startups like humanoid robot maker FigureAI and 1X Technologies. OpenAI CEO Sam Altman has even gotten into the game himself by investing in medical AI startups like Thrive AI Health.
Another notable move is OpenAI’s partnership with Apple, which aims to bring ChatGPT to a wider audience through Apple products. While the details of the deal have yet to be announced, the partnership shows the huge potential of AI to change the way we interact with technology.
While it’s no secret that companies are pouring money into AI, the actual spending is still a mystery. It’s estimated that around $1 trillion will be poured into the field in the near future. What’s remarkable is that while tech giants publicly disclose their investments in startups, they’re remarkably tight-lipped about what’s going on behind the scenes, from buying chips to hiring experts and building data centers.
Google has positioned itself as an AI leader and has released a number of products in recent months, although some have been recalled due to inaccuracies. However, executives remained tight-lipped about the company’s AI spending during its second-quarter earnings call.
Executives have not provided specific numbers on the company’s AI investments, but continue to emphasize that its AI ambitions will pay off in the long term. For now, what we know is that Google has spent $3 billion building and expanding its data centers, and $60 million on AI training.
OpenAI
OpenAI’s operating expenses could reach $8.5 billion this year, according to The Information, a figure that could be more than the startup can afford, raising concerns about its profitability, especially as competitors offer free versions of their chatbots.
The costs include about $4 billion to rent servers from Microsoft, $3 billion to train AI models with new data, and $1.5 billion in labor costs, the report said.
Meta
Meta CEO Mark Zuckerberg said he plans to buy 350,000 graphics processing units (GPUs) from semiconductor maker Nvidia by the end of 2024, bringing Meta's total GPU count to about 600,000. Analysts estimate the giant could have spent about $18 billion by the end of 2024.
An analyst at the world's largest financial services group JPMorgan (USA) said the company's expenses are driven by spending on AI, which could reach $50 billion by 2025.
Apple
During the first-quarter 2024 earnings call, Apple CFO Luca Maestri said the company has spent about $100 billion over the past five years on technology research and development, but did not specify how much of that was for AI. Maestri said the company will continue to invest in some areas and partner with third parties in others.
Microsoft
Documents leaked earlier this year revealed the company's plans to have 1.8 million AI chips by the end of the year and boost its data center capacity.
Microsoft also said it would continue to invest in AI and cloud services during its third-quarter earnings call in April. Earlier, sources said Microsoft had invested $13 billion in artificial intelligence startup OpenAI.
Amazon
Amazon is developing an AI chatbot to compete directly with OpenAI's ChatGPT, called Metis, and has attempted to create its own AI chip to compete with Nvidia, though with limited success.
Meanwhile, the company plans to spend nearly $150 billion over the next 15 years on data centers. Amazon also plans to invest up to $230 million in startups building applications powered by generative AI.
About $80 million of the investment will fund Amazon's AWS Generative AI Accelerator program, an initiative by Amazon Web Services (AWS) to support and accelerate the growth of startups working in the field of generative AI.
Much of the new investment is in the form of compute credits for AWS infrastructure, meaning it cannot be transferred to other cloud providers like Google and Microsoft.
xAI
Earlier this month, billionaire Elon Musk, the boss of xAI startup, said that the latest version of xAI's Grok chatbot will be trained on 100,000 H100 GPU chips. Nvidia's H100 GPU chips help process data for large language models and are a key component in scaling AI.
Each GPU is estimated to cost between $30,000 and $40,000, meaning xAI is spending between $3 and $4 billion on AI chips.
It is worth noting that it is unclear whether Elon Musk's company will buy those chips outright. Previously, The Information reported that xAI was in discussions with the world's leading multinational technology corporation, famous for its software solutions and cloud services, Oracle (USA), about investing $ 10 billion to rent cloud servers.
Instead of making a fixed investment in purchasing a bunch of H100 GPU chips, xAI may be considering options for renting cloud servers from providers like Oracle. This would allow the company to optimize costs and focus on product development while participating in the increasingly fierce AI race.