Oracle Brings Google Gemini Models to Enterprise Customers
The partnership aims to give users more choice when building AI agents and automating business processes.
Google Gemini
Alphabet fell 5%, its worst day in a year, after two top researchers quit for OpenAI and Anthropic
Alphabet’s stock closed down almost 5% on Monday, making it Google’s worst trading day in over a year.
The decline coincided with two prominent researchers leaving to join competitors and growing concerns about artificial intelligence.
It was the company’s biggest decline since a roughly 7% decline in May 2025, and it was larger than the Nasdaq and the other major tech names.
A series of employee layoffs at Google’s primary AI teams preceded the sell-off.
Noam Shazeer, a vice president of engineering and co-lead of Google’s Gemini AI models, announced on Wednesday that he was quitting to join rival OpenAI, which sparked the controversy last week.
Less than two years prior, Shazeer had returned to Google.
John Jumper, a Nobel Prize winner and senior research scientist at Google, announced late on Friday that he was quitting the DeepMind AI lab to join Anthropic.
Analysts speculate that Google may be losing ground in the AI battle and finding it difficult to retain top AI personnel as a result of these exits.
Last year, Google briefly took the lead with an innovative methodology, but since then, rivals and well-funded AI firms have escalated the competition for talent.
Concerns regarding Google’s position have been heightened by the departure of important AI researchers to competitors.
At just 13% odds, traders are placing bets on Polymarket to determine which AI model will be the best by the end of July.
Search business under threat
Since the introduction of ChatGPT, generative AI has posed a danger to Google’s future. Recently, ChatGPT surpassed one billion monthly active users.
There are two risks: first, Google would lose its dominance; second, by retaliating, it might harm its own search business, which depends primarily on advertising. There are some visible cracks.
While ChatGPT traffic has increased over the last month, Google’s search traffic has decreased by more than 1%.
As it promotes a “no-AI” option through new browser features, search engine DuckDuckGo is experiencing an increase in installs.
Open-source Chinese models like DeepSeek and z.AI may also put pressure on Google and other leading US AI companies.
These are significantly less expensive options and have capabilities comparable to those of the major American versions.
Spending fears rattle investors
In addition to the loss of expertise, concern over the enormous expense of developing AI is growing.
Investors are blaming the AI trade for the global decline in markets.
Depending on how much risk investors are willing to take, the boom has been powered by billions of dollars being invested in software and data centers.
In fiscal 2026, Google alone has announced intentions to spend between $180 billion and $190 billion, primarily on data centers and AI computing.
It is now difficult to disregard the financial math. Alphabet has cautioned that its capital expenditures will increase significantly in 2027 and anticipates spending between $180 and $190 billion in 2026.
Operating cash flow increased to $45.8 billion in the first quarter, but capital expenditures more than doubled to $35.7 billion.
This reduced free cash flow from over $19 billion a year earlier to little over $10 billion.
The tech industry’s poor day reflects rising skepticism about whether all of this expenditure will result in long-term profit.
“Simply said, the market is drawing a sharp line between AI spenders and AI earners,” Wagner stated.
“The big spending hurts the hyperscalers’ margins, while those massive hardware orders directly benefit memory manufacturers.”
In a recent Wall Street Journal interview, Microsoft CEO Satya Nadella stated that models may become more affordable and simpler to replace, raising questions among investors about whether the expenditure actually provides a competitive advantage.
The uncertainty permeated the tech industry.
Dutch semiconductor manufacturer ASML had a 5% decline, while memory chip manufacturers Samsung Electronics and SK Hynix each saw a 12% decline.
After dropping 16% on Monday, SpaceX shares appeared likely to continue their three-day losing skid.
There were also additional losses for the larger Magnificent Seven group.
Don’t just read crypto news. Understand it. Subscribe to our newsletter. It’s free.
Japan’s manufacturers see factory floors as AI’s next frontier
Japan is searching for its place in the global AI race. While American and Chinese companies dominate AI models and computing infrastructure, Japanese companies believe their expertise in robotics could help pioneer AI in real world tasks.
On May 13 Japanese industrial equipment maker, Fanuc, announced a partnership with Google that aims to create factory robots that can understand spoken and handwritten instructions and carry out factory tasks autonomously.
Fanuc, founded in Japan in 1956, is one of the world’s largest industrial machinery manufacturers. It’s developed an AI system with the help of Google Gemini that can be operated without programming skills. It plans to make all its robots compatible with Google software.
In December 2025, Fanuc also announced a collaboration with NVIDIA which will see it open its previously closed robot software systems. At a press conference on May 13, Senior Managing Officer Kenishiro Abe said the partnership stems from the limitations of developing an entire AI ecosystem in-house. It plans to incorporate AI systems from a host of different companies.
What is physical AI?
Factories are set to benefit the most from physical AI. While robots are already used extensively, they remain limited to repetitive tasks.
Physical AI is the practical application of AI. These AI systems are trained to perceive the real world, reason with it, act autonomously in real time as well as learn and collaborate from humans. They excel at handling complex and unpredictable tasks.
What is the winning formula?
For decades, Japanese factories have been shaped by knowledge that was never written down. Now, Japanese companies are trying to teach that knowledge to machines.
According to a Nomura Securities report, Japan’s decades-long manufacturing expertise and factory-floor data could power industrial humanoid robots.
In the 1990s Japanese manufacturers made up 80 percent of the global industrial robot market, according to the International Federation of Robotics. The figure has since fallen to roughly 40 percent.
As of 2024, Chinese companies such as Estun Automation and Inovance Technology are gaining ground and account for 40 percent of the global humanoid robot market.
But many Chinese companies still rely on Japanese machinery components. Nomura Securities predicts that Japan’s expertise in motion control technologies, industrial datasets, precision manipulators (i.e. robot hands) and semiconductor equipment could drive growth in a post-2030 economy.
Japan’s gaping ‘digital deficit’
Fanuc’s decision to open its source robotics software is a significant pivot from the Japanese manufacturing sector’s emphasis on hardware.
The country trails behind the U.S. and China in AI digital transformation (DX). Japanese companies rely heavily on software from U.S. tech giants resulting in a massive ‘digital deficit’ in which payments for digital services flow overseas.
The Ministry of Economy, Trade and Industry (METI) recorded a $4.9 billion digital services deficit in 2023. The U.S. on the other hand, posted a $173.7 billion surplus while China logged a $40.4 billion digital surplus.
As companies integrate AI into manufacturing, the Japanese government anticipates that rising demand for industrial robots will support the growth of Japanese industrial machinery companies.
Japanese technology company ARUM Inc has developed a fully automated, AI-enabled production line for metal part manufacturers. Its TTMC system costs approx. $2.3 million each. At Tokyo Sushi Tech Expo 2026, the company said it will install 100 units across Japan and has received enquiries from South Korea and the United States.
“We are not simply selling machines. We are connecting them through the cloud and building infrastructure,” said Takayuki Hirayama, CEO of ARUM Inc.
ARUM Inc believes that AI-driven manufacturing automation can solve global labor shortages and changing career preferences.
“Even in younger countries like India and Southeast Asia, skilled manufacturing workers are disappearing because IT and tourism are seen as more lucrative.”
The Japanese government wants to lead the robotics AI race
At a New Years press conference, Japanese Prime Minister Sanae Takaichi announced plans to accelerate physical AI innovation and expand the technology globally. She stated that AI-powered robots will learn from high-quality domestic data, in particular, Japan’s long established factory know-how.
The initiative builds on remarks made in December 2025 when Takaichi directed the government to support domestically produced general purpose AI models which are an essential component of physical AI. METI is set to launch a one trillion yen funding package (approx. $6.45 billion) over five years to help develop Japanese physical AI.
CEO Masato Fujino of Japanese industrial devices company, Fairy Devices Inc, believes that the challenge is no longer using AI within computers but bringing AI into the real world.
The company has produced wearable AI devices that prevent technicians from missing important checks. They are built with cameras, microphones, sensors and communications capabilities. The devices have accumulated large volumes of data and have trained the company’s vision language model which aims to replace experts such as air conditioner repair technicians.
At Tokyo Sushi Tech Expo 2026, Fujino said specialized data directly from skilled workers is indispensable for industrial AI systems.
“Google Gemini is powerful because Google owns Youtube. But when it comes to highly specialized industrial tasks, such as repairing industrial equipment, that data does not exist on Youtube.”
What role can Japan play in the physical AI sector?
Japan’s answer to AI is not frontier models but industrial data. Despite fierce competition for low-cost, high-quality physical AI, Japanese industry leaders are optimistic about Japan’s trajectory.
In their eyes Japan’s reputation for manufacturing excellence and proven track record in factory automation is difficult to replicate anywhere else.
If you’re reading this, you’re already ahead. Stay there with our newsletter.