Chinese tech giants are training their models offshore to sidestep US curbs on Nvidia’s chips
Nvidia can’t sell its best AI chips in the world’s second largest economy. That’s an Nvidia problem. But it’s also a China problem — and it’s one that the region’s tech giants have resorted to solving by training their AI models overseas, according to a new report from the Financial Times.
Citing two people with direct knowledge of the matter, the FT reported that “Alibaba and ByteDance are among the tech groups training their latest large language models in data centers across south-east Asia.” Clusters of data centers have particularly boomed in Singapore and Malaysia, with many of the sites kitted out with Nvidia’s latest architecture.
One exception, per the FT, is DeepSeek, which continues to be trained domestically, having reportedly built up a stockpile of Nvidia chips before the US export ban came into effect.
Last week, Nvidia spiked on the news that the Trump administration was reportedly considering letting the tech giant sell its best Hopper chips — the generation of chips that preceded Blackwell — to China.
Citing two people with direct knowledge of the matter, the FT reported that “Alibaba and ByteDance are among the tech groups training their latest large language models in data centers across south-east Asia.” Clusters of data centers have particularly boomed in Singapore and Malaysia, with many of the sites kitted out with Nvidia’s latest architecture.
One exception, per the FT, is DeepSeek, which continues to be trained domestically, having reportedly built up a stockpile of Nvidia chips before the US export ban came into effect.
Last week, Nvidia spiked on the news that the Trump administration was reportedly considering letting the tech giant sell its best Hopper chips — the generation of chips that preceded Blackwell — to China.
Alibaba unveils its first AI glasses, taking on Meta directly in the wearables race
Retail and tech giant Alibaba launched its first consumer-ready, AI-powered smart glasses on Thursday, marking its entrance into the growing wearables market.
Announced back in July, the Quark AI glasses just went on sale in the Chinese retailer’s home market, with two versions currently available: the S1, starting at 3,799 Chinese yuan (~$536), and the G1 at 1,899 yuan (~$268) — a considerably lower price than Meta’s $799 Ray-Ban Display glasses, released in September.
Musk: Tesla’s Austin Robotaxi fleet to “roughly double” next month, but falls well short of earlier goals
Yesterday, Elon Musk jumped onto a frustrated user’s post on X, who was complaining that they were unable to book a Robotaxi ride in Austin. Musk aimed to reassure the would-be customer that the company was expanding service in the city:
“The Tesla Robotaxi fleet in Austin should roughly double next month,” wrote Musk.
While that sounds impressive, there are reports that Austin only has 29 Robotaxis in service.
But last month, Musk said the Robotaxi goal was to have “probably 500 or more in the greater Austin area,” by the end of the year.
Meanwhile, Google’s Waymo has more than 100 autonomous taxis running in Austin, and 1,000 more in the San Francisco Bay Area.
“The Tesla Robotaxi fleet in Austin should roughly double next month,” wrote Musk.
While that sounds impressive, there are reports that Austin only has 29 Robotaxis in service.
But last month, Musk said the Robotaxi goal was to have “probably 500 or more in the greater Austin area,” by the end of the year.
Meanwhile, Google’s Waymo has more than 100 autonomous taxis running in Austin, and 1,000 more in the San Francisco Bay Area.
Apple to beat Samsung in smartphone shipments for first time in 14 years
Thanks to Apple’s popular iPhone 17, the company is on track to ship more smartphones than rival Samsung for the first time in 14 years, according to a report from CNBC.
Counterpoint Research projects that Apple will ship about 243 million phones to retailers this year, capturing 19.4% of the global market.
Samsung will come in just behind Apple, with 235 million phones shipped, giving it an 18.7% global market share, per the report.
A favorable upgrade cycle, plus an expected lower-cost entry-level iPhone next year, are among the factors expected to keep Apple in the lead for the next few years.
Samsung will come in just behind Apple, with 235 million phones shipped, giving it an 18.7% global market share, per the report.
A favorable upgrade cycle, plus an expected lower-cost entry-level iPhone next year, are among the factors expected to keep Apple in the lead for the next few years.
OpenAI eyes 220 million paid subscribers by 2030, The Information reports
OpenAI is scrambling to figure out how to generate enough steady revenue to turn the expensive AI services it offers into profits, as it spends dizzying sums on the infrastructure needed to scale its business to the expected demand.
It appears that for now, the company's solution comes straight from the old, reliable Big Tech playbook: turn free users into paying subscribers.
According to The Information, OpenAI is projecting that it can ramp up to about 220 million paid subscribers by 2030.
The company currently has about 800 million users, with 35 million of them paying for Plus or Pro subscriptions, at either $20 or $200 per month, per the report. OpenAI thinks that in five years it will have 8.5% of its projected 2.6 billion weekly active users paying for a Plus plan, or about 220 million people, The Information reports.
That would put ChatGPT in the same league as Spotify (281 million subscribers in September 2025), and Netflix (302 million subscribers in December 2024).
According to The Information, OpenAI is projecting that it can ramp up to about 220 million paid subscribers by 2030.
The company currently has about 800 million users, with 35 million of them paying for Plus or Pro subscriptions, at either $20 or $200 per month, per the report. OpenAI thinks that in five years it will have 8.5% of its projected 2.6 billion weekly active users paying for a Plus plan, or about 220 million people, The Information reports.
That would put ChatGPT in the same league as Spotify (281 million subscribers in September 2025), and Netflix (302 million subscribers in December 2024).
Bitcoin’s plunge could hurt Tesla’s bottom line
Sometimes bitcoin giveth, but lately it’s been taking away from Tesla.
A new accounting rule that took effect earlier this year requires Tesla to include unrealized gains and losses on its bitcoin holdings in its quarterly results. According to analyst Troy Teslike, Tesla is facing an unrealized loss of more than $300 million in the fourth quarter on its 11,509 bitcoin, thanks to bitcoin’s recent plunge. That would reduce its GAAP earnings per share by about $0.10. If bitcoin plummets further, say to $60,000, that unrealized loss could grow to more than $600 million, with a -$0.19 impact on EPS.
For context, the FactSet analyst consensus for Tesla’s net income in Q4 is penciled in at $1.6 billion with GAAP EPS of $0.37, so additional losses would represent a big earnings headwind. For a company already navigating margin pressure, bitcoin’s volatility adds one more wild card to the mix.
Elon Musk: We’ve tried to license Tesla’s FSD technology to legacy automakers but “they don’t want it”
Tesla CEO Elon Musk has repeatedly said his company is open to licensing its Full Self-Driving technology to major automakers so that they could potentially make their own fleets drive themselves. Now, Musk is saying that those automakers aren’t interested.
“I’ve tried to warn them and even offered to license Tesla FSD, but they don’t want it! ” Musk posted on X.
I’ve tried to warn them and even offered to license Tesla FSD, but they don’t want it! Crazy …
— Elon Musk (@elonmusk) November 24, 2025
When legacy auto does occasionally reach out, they tepidly discuss implementing FSD for a tiny program in 5 years with unworkable requirements for Tesla, so pointless. 🤷♂️
🦕 🦕
While the post is presumably meant to convey that the auto industry is out of touch and behind the times, it also suggests an anticipated future revenue source for Tesla so far isn’t panning out.
I’ve tried to warn them and even offered to license Tesla FSD, but they don’t want it! Crazy …
— Elon Musk (@elonmusk) November 24, 2025
When legacy auto does occasionally reach out, they tepidly discuss implementing FSD for a tiny program in 5 years with unworkable requirements for Tesla, so pointless. 🤷♂️
🦕 🦕
While the post is presumably meant to convey that the auto industry is out of touch and behind the times, it also suggests an anticipated future revenue source for Tesla so far isn’t panning out.
Apple cuts sales jobs in rare layoff
Apple is cutting “dozens” of roles from its sales team in a rare layoff, according to a report from Bloomberg. The reductions are aimed at streamlining the company’s sales to businesses, schools, and government accounts, per the report.
Apple rarely turns to layoffs, compared to its tech peers, making the reduction noteworthy.
An Apple spokesperson told Bloomberg: “To connect with even more customers, we are making some changes in our sales team that affect a small number of roles,” and that the employees will be able to apply for new roles in the company.
An Apple spokesperson told Bloomberg: “To connect with even more customers, we are making some changes in our sales team that affect a small number of roles,” and that the employees will be able to apply for new roles in the company.
Anthropic releases Claude Opus 4.5 as AI war heats up
The past few weeks have seen new, impressive AI models debut from OpenAI and Google. Today it’s Anthropic’s turn to flex, as it releases Claude Opus 4.5, the latest iteration of its flagship AI model.
Anthropic’s Claude model is widely considered to be among the best at coding, and this model helps the company stay at the head of the pack.
Benchmarks released by Anthropic show Opus 4.5 besting both GPT-5.1 and Gemini 3 with an all-time high score of 80% and the widely used SWE-bench coding benchmark. It also posted high scores for benchmarks measuring computer use and the notoriously challenging ARC-AGI-2 visual problem-solving test, though apparently it can’t run a vending machine as profitably as Google’s Gemini 3 can.
AI coding is one of the few bright spots as companies seek profitable enterprise applications for AI that actually improve productivity. Anthropic’s success with enterprise customers has helped push its valuation to nearly $350 billion.
Benchmarks released by Anthropic show Opus 4.5 besting both GPT-5.1 and Gemini 3 with an all-time high score of 80% and the widely used SWE-bench coding benchmark. It also posted high scores for benchmarks measuring computer use and the notoriously challenging ARC-AGI-2 visual problem-solving test, though apparently it can’t run a vending machine as profitably as Google’s Gemini 3 can.
AI coding is one of the few bright spots as companies seek profitable enterprise applications for AI that actually improve productivity. Anthropic’s success with enterprise customers has helped push its valuation to nearly $350 billion.
Amazon plans to invest up to $50 billion for “AI and supercomputing infrastructure” for the US government
Amazon said it will invest up to $50 billion to build out its AI computing infrastructure for the US government.
Based on the company’s AWS cloud platform, Amazon will help build up to 1.3 gigawatts of dedicated AI high-performance computing infrastructure, according to a press release announcing the plans.
The project, which will including building new data centers, is set to break ground in 2026.
Amazon AWS CEO Matt Garman said:
“We’re giving agencies expanded access to advanced AI capabilities that will enable them to accelerate critical missions from cybersecurity to drug discovery. This investment removes the technology barriers that have held government back and further positions America to lead in the AI era.”
The new computing capacity will be available to agencies through use of AWS government products AWS Top Secret, AWS Secret, and GovCloud Regions.
The project, which will including building new data centers, is set to break ground in 2026.
Amazon AWS CEO Matt Garman said:
“We’re giving agencies expanded access to advanced AI capabilities that will enable them to accelerate critical missions from cybersecurity to drug discovery. This investment removes the technology barriers that have held government back and further positions America to lead in the AI era.”
The new computing capacity will be available to agencies through use of AWS government products AWS Top Secret, AWS Secret, and GovCloud Regions.
Amazon now has 900 data centers spread across 50 countries, report says
The exact size and shape of Amazon’s AWS global network of data centers has always been a closely guarded secret. A new report from Bloomberg and SourceMaterial sheds some light on AWS’s global reach.
Based on internal documents seen by Bloomberg, Amazon’s cloud operations include more than 900 data centers spread across 50 countries.
Amazon owns the majority of its data centers, but contracts with at least 180 different colocation entities, according to the report.
Musk: We will build AI chips at higher volumes “than all other AI chips combined”
In a late-night post on X, Elon Musk boasted about Tesla’s custom AI chip plans.
Musk said the current version of Tesla’s AI chip, the A14, is in cars and data centers today, while work is underway on the A15 and A16. The goal is an Apple-style yearly iteration of its workhorse custom AI chip.
Musk wrote: “We expect to build chips at higher volumes ultimately than all other AI chips combined. Read that sentence again, as I’m not kidding.”
Even OpenAI is worried about Google’s Gemini 3
When OpenAI’s ChatGPT burst onto the scene in November 2022, it sent shock waves through Silicon Valley’s biggest names. Google, Microsoft, and Amazon had all been developing generative AI, but OpenAI’s breakthrough sparked an all-out race to catch up. Until now.
It seems that OpenAI CEO Sam Altman is feeling the heat from Google, whose newly released Gemini 3 has been receiving stellar reception from AI leaderboards, analysts, and consumers alike.
“We know we have some work to do but we are catching up fast,” OpenAI CEO Sam Altman told colleagues last month, after learning about Google’s AI advances, The Information reports. “I expect the vibes out there to be rough for a bit.”
Google’s AI progress, Altman said, could “create some temporary economic headwinds for our company,” but he said OpenAI would emerge on top.
However, it’s worth remembering that, despite OpenAI’s first-mover advantage and supersized valuation, Google is a substantial adversary that is peppering its AI models across its giant existing — and highly lucrative — product suite.
It seems that OpenAI CEO Sam Altman is feeling the heat from Google, whose newly released Gemini 3 has been receiving stellar reception from AI leaderboards, analysts, and consumers alike.
“We know we have some work to do but we are catching up fast,” OpenAI CEO Sam Altman told colleagues last month, after learning about Google’s AI advances, The Information reports. “I expect the vibes out there to be rough for a bit.”
Google’s AI progress, Altman said, could “create some temporary economic headwinds for our company,” but he said OpenAI would emerge on top.
However, it’s worth remembering that, despite OpenAI’s first-mover advantage and supersized valuation, Google is a substantial adversary that is peppering its AI models across its giant existing — and highly lucrative — product suite.
