Wayne Ma
Verified
As seen in:
SemiAnalysis,
Fox News,
The Wall Street Journal,
MarketWatch,
SemiAnalysis (Newsletter),
Flipboard,
Fox Business,
Nasdaq,
DEV Community,
The Information,
Forte.ee
and
I'm an analyst @SemiAnalysis_, formerly with The Information and The Wall Street Journal in Singapore, Beijing and Hong Kong.
Is this you? As a journalist, you can create a free Muck Rack account to customize your profile, list your contact preferences, and upload a portfolio of your best work.
Claim your profile
Articles by Wayne Ma
Stop AI agents from reviving old project decisions
Long-running repositories collect Markdown files that disagree about what is current. Once Codex, Claude, and humans work across many sessions, an old architecture note can quietly become "the plan" again. Markdown Gatekeeper keeps one current revision per topic. Competing docs stay proposals; superseded sources remain recoverable. Local-first. Plain Markdown and Git. No account or cloud service. npm install -g github:nanlogic/markdown-gatekeeper mdg setup codex mdg init .
Your AI Can Write the Sales Email. Can You Trust It to Click Send?
Ask AI to write a sales email and it can finish before you take a sip of coffee. Ask it to open the right account, update the CRM, attach the right file, change a status, click Save, and confirm the change—and the babysitting begins. Every browser click becomes a tiny trust exercise. Is this the right customer? Is the page still current? Did the button move after a rerender? Did Save actually work? If the connection dropped after the click, will the Agent do it twice? This is the browser gap.
Meta’s Infrastructure Team Needs A Culture Reset
In our recent newsletter piece about Meta Superintelligence, we expressed reasons to be optimistic on Meta AI. MSL now has many of the right ingredients to catch up with Anthropic and OpenAI to return to the frontier. However, we also briefly alluded to cultural issues plaguing Meta’s infrastructure teams. This article will dive into how these cultural issues have manifested into expensive missteps, whether it be with acquisitions like Rivos or strange choices on hardware architecture.
AI Agents Hallucinate Recommendations. Here's the Fix....
My agent recommended Supabase for a side project. It had never connected to Supabase, never run a query, never hit a rate limit. It said Supabase because other agents said Supabase. Pattern matching filed it under "safe recommendation." The project broke three weeks in on an auth edge case. I spent a weekend fixing it. Your Claude instance recommends Stripe the same way. It echoes training data from months ago. Cursor suggests Vercel for deployment without having deployed anything.
Brain-Inspired Memory for LLMs
Your brain does three things with memory that LLMs don't: it forgets what's irrelevant, it connects related ideas when you recall one, and it consolidates fragmented experiences into coherent knowledge while you sleep. I borrowed all three for nan-forget, a long-term memory system I built with Claude Code for AI coding tools. The Problem I use Claude Code daily.
Nvidia Needed Groq After All
Greetings from the San Jose Convention Center, where Nvidia is throwing its annual AI conference. The AI craze is palpable here. Humanoid robots offer fist-bumps to conferencegoers, and attendees are buying sweaters with Nvidia CEO Jensen Huang’s face on them. Nebius, a cloud provider that rents out Nvidia graphics processing units, took over a bar and served guests server-themed cocktails, such as a GPU Spritz and Ryes of the Machines.
Nvidia Unveils NemoClaw Agent Software
Source: The Information Nvidia CEO Jensen Huang said Monday the company will offer new software tools that allow customers to build their own open-source AI agents for business use. Launch of the software, NemoClaw, follows the popularity of OpenClaw, open-source software to create personal agents that can write code, edit files and surf the web on behalf of users.
Nvidia Unveils Groq-Based Chip System to Speed Up AI Tasks Like Coding
Source: The Information Nvidia CEO Jensen Huang announced a new AI server system based on technology the company licensed last year from chip startup Groq, designed to make its AI servers more energy- and cost-efficient for tasks such as AI coding. This is the first time Nvidia has integrated another company’s AI processor into its server racks, and it’s a tacit admission that its graphics processing units, on their own, may not always be the best option for certain AI workloads.
Shell looks to boost investment in downstream China gas sector
By WAYNE MA BEIJING -- Royal Dutch Shell hopes to boost investment not only upstream but also in the downstream gas sector in China, as it wants to profit from every link in the gas supply chain, the head of the company's China unit said Monday. I don't want to just be the largest international oil company involved in LNG supply to China, Lim Haw Kuang, Shell's executive chairman in China, told Dow Jones Newswires.
Nvidia to Launch Platform for OpenClaw-Like AI Agents
Source: Wired Nvidia is planning to launch a platform for AI agents aimed at enterprise customers called NemoClaw, Wired reported Monday. The platform will be open-source, it said, meaning the software code of these AI agents will be publicly available and modifiable. Nvidia already has reached out to Salesforce, Cisco, Google, Adobe and Crowdstrike over potential partnerships, though it isn’t clear whether this resulted in actual deals, Wired reported.
Apple Discusses Google Hosting New Siri as Need for Cloud Help Grows
In January, when Apple announced an agreement to use Google’s Gemini models in Apple products, it was an admission that the iPhone maker hasn’t been able to compete in AI on its own over the past couple years. But the deal also showed that Apple is again relying on an outside company to help it with another deficiency: in cloud computing. Now there’s a chance Apple could deepen its cloud dependence on Google even more.
Nvidia to Announce New Chip With Groq’s Technology
Source: The Wall Street Journal Nvidia is expected to announce a new chip later this month using technology from startup Groq, The Wall Street Journal reported late Friday. The chip is aimed at speeding up inference computing, the term for running AI on servers after AI models have been fully developed, the newspaper said. The industry’s demand for AI chips is gradually shifting from training AI models to operating the models for users.
Meta’s Internal Chip Design Efforts Hit Roadblocks
As Meta Platforms strikes new chip supply deals with AMD and Nvidia, it has been running into problems with AI chips it is designing internally, according to six people with direct knowledge of the matter. Meta last week scrapped the most advanced chip it was developing for training AI models, after struggling with the chip’s design, and shifted its focus to a less complicated version, the people said.
Exclusive: OpenAI Hires Meta AI Researcher Who Previously Led Apple’s Models Team
Source: The Information OpenAI has hired a high-profile AI researcher at Meta Platforms who had joined the company from Apple about seven months ago, a spokesperson from OpenAI confirmed. Ruoming Pang, who led AI infrastructure for Meta Superintelligence Labs, which is developing new AI models, left the company last week and has joined OpenAI.
Meta Strikes Six-Gigawatt Compute Deal with AMD
Source: The Information Meta Platforms struck a deal to buy enough AMD’s AI chips to power six gigawatts of data center capacity over several years, the companies announced Tuesday. In exchange, Meta stands to get 10% of AMD, currently worth $34 billion, if it buys all of the chips. Six gigawatts is an enormous amount of capacity: for context, a traditional data center typically consumed up to 50 megawatts of power. The deal is a big legup for AMD, which badly trails Nvidia in the AI chip market.
Nvidia, MediaTek to Launch Chip for Consumer PCs This Year
Source: The Wall Street Journal Nvidia is planning to release a new chip in the first half of this year for consumer laptops and desktops in collaboration with Taiwanese chip designer MediaTek, The Wall Street Journal reported. PC makers such as Dell Technologies and Lenovo Group are working on products using the chip, which has both a central processing unit and a graphics processing unit on a single piece of silicon, the newspaper said.
Disney, Paramount Send Legal Letters to ByteDance Over AI Videos
Source: Variety Disney, Paramount Skydance and the largest union for U.S. entertainment professionals sent cease-and-desist letters to ByteDance after videos generated by the Chinese tech giant’s AI video model, Seedance 2.0, went viral on social media, the entertainment publication Variety reported.
Hollywood Group Slams ByteDance’s AI Video Service for Copyright Breach
Source: The Information The Motion Picture Association, which represents Hollywood studios, has called on ByteDance to “cease its infringing activity” after videos generated by the Chinese tech giant’s new AI video model, Seedance 2.0, went viral on social media. “In a single day, the Chinese AI service Seedance 2.0 has engaged in unauthorized use of U.S. copyrighted works on a massive scale,” said Charles Rivkin, chairman and CEO of the MPA, in a statement.
U.S. Lawmakers Urge Trump to Close Chip-Equipment Loopholes
Source: The Information The chairs of the U.S. House China and Foreign Affairs committees sent a letter urging the Trump administration to close loopholes in export controls that allow China to gain access to advanced chip-making equipment. The push comes after media reports in December that China made progress developing advanced chip-making machines, either by upgrading older equipment or by building prototypes using scavenged parts.
Nvidia to Delay New Gaming Chip Due to Memory Chip Shortage
Nvidia won’t release a new graphics chip for gamers this year due to a deepening global shortage of memory chips, prompted by the AI boom, according to two people with direct knowledge of the matter. It would be the first year in three decades that Nvidia hasn’t released a new graphics processing unit for gaming. Nvidia got its start in the early 1990s by designing graphics chips for videogames and consoles before it expanded into high-performance AI computing chips in the early 2010s.
Nvidia’s H200 Chip Sales to China Remain Stalled By U.S. Government
Source: Financial Times Nvidia’s application to sell the H200 chips to China is still awaiting final approval by the U.S. government, nearly two months after President Donald Trump approved the sale, the Financial Times reported, citing people familiar with the discussions. The stall highlights how Nvidia is increasingly caught between the U.S. and China as both sides seek control of their AI chip sales to advance their own national security and economic interests.
With Google Deal, Apple’s Craig Federighi Plots a Cautious Course in AI
Last fall, Apple’s software chief, Craig Federighi, delivered a warm message to a meeting of the company’s combined software and AI groups. Federighi said he was excited about deeper collaboration between his software team and Apple’s AI group. But then came Federighi’s criticism. He added that he liked moving fast and feeling the “wind at his back,” but in recent years he had felt Apple hadn’t been moving as quickly as he would have liked in AI.
Apple Developing AI Wearable Pin in Challenge to OpenAI, Meta
Source: The Information Apple is developing an AI-powered wearable pin that could be released as early as 2027, according to people with direct knowledge of the matter. The device would position Apple to compete more effectively with OpenAI, which is planning its own AI-powered devices, and Meta Platforms, which is already selling smart glasses.
Nvidia Taps Google Veteran as Its First Chief Marketing Officer
Source: The Information Nvidia is hiring its first chief marketing officer who will report directly to CEO Jensen Huang, an Nvidia spokesperson said. Alison Wagonfeld, who previously headed marketing at Google Cloud, announced the move on Thursday on LinkedIn. Wagonfeld spent nearly a decade at Google and will join Huang’s leadership team to spearhead marketing and communications, she wrote. The spokesperson said she would begin in February.
Nvidia Unveils Rubin and Announces New Open Source Autonomous Driving Model
Source: The Information Nvidia CEO Jensen Huang on Monday announced a new autonomous driving model that aims to help cars reach Level 4 autonomous driving, in which a driver isn’t required to supervise the vehicle. Separately, Huang also announced more detailed specifications for the company’s upcoming next-generation flagship AI chips, known as Rubin.
Nvidia’s Big Ambitions to Solve Manufacturing Shows Slow Returns So Far
In the past two years, Nvidia’s business of selling chips for AI has rocketed into the stratosphere, lifting the company’s revenue to nearly $148 billion in the nine months through October, up from $27.5 billion for the same period in 2023. But CEO Jensen Huang isn’t satisfied.
Inside Apple’s iPhone Road Map, From Foldable Screens to Curved Glass Cases
The iPhone family is growing again. Over the next two years, Apple is planning a major expansion of its smartphone product line that could see it selling at least seven new models of its flagship device by fall 2027, up from five iPhone models today, according to multiple people working directly on the products at both Apple and its suppliers.
The People Running Apple After Its Executive Exodus
For years, Apple was known for having very little turnover in its top ranks. Not anymore. In recent weeks, the company has seen a stream of high-profile executives head for the exits, at a moment when Apple is scrambling to build the AI features consumers increasingly expect from the iPhone.
Netflix in Exclusive Talks with Warner Bros.
Source: Bloomberg Warner Bros. Discovery has entered into exclusive talks to sell its film and TV divisions and streaming service to Netflix, Bloomberg reported. The move would represent a seismic shift in the media landscape by combining the world’s most popular streaming service with one of Hollywood’s most prominent studios. Bloomberg said that Netflix is offering a $5 billion breakup fee if regulators don’t approve the deal and said it could be announced in the coming days.
Nvidia Says Its Server Boosts Performance of Top Open-Source AI Models
Source: The Information Nvidia said one of its most powerful AI servers, which enables 72 chips to work in unison, can boost the performance of some of the world’s top open-source artificial intelligence models by a factor of 10. The announcement comes as Nvidia faces intense competition from other AI chips including Google’s tensor processing unit and Amazon’s Tranium chips.
Analyst: Apple to Use Intel to Manufacture Low-End Chips for Macs and iPads
Source: The Information Apple plans to use Intel to manufacture low-end chips for its laptops and iPads, according to Ming-Chi Kuo, a prominent Apple supply chain analyst.
Intel CEO Extends Support for Senior Executive After TSMC lawsuit
Source: The Information Intel CEO Lip-Bu Tan said Wednesday in an internal memo to staff that he fully supports the executive who’s been sued by Taiwan Semiconductor Manufacturing Co. over alleged breach of trade secrets, and that there’s no merit to the allegations. The U.S. company only began hiring discussions with Wei-Jen Lo after he retired earlier this year from TSMC, Tan said in the memo. He was responding to a lawsuit filed by TSMC on Tuesday accusing Lo of leaking trade secrets to Intel.
Tenstorrent Nears $3 Billion Valuation in Fidelity-Led Round
Chip startup Tenstorrent has been in talks to raise at least $800 million in a deal led by current investor Fidelity Management, according to two people with direct knowledge of the talks. The investment would value the Toronto-based firm at around $3.2 billion before the investment, a 60% increase from its valuation not including the investment last year, one of the people said.
Nvidia CEO Warns China Will Win AI Race
Source: The Financial Times Nvidia CEO Jensen Huang said China will win the artificial intelligence race because of the country’s lower energy costs and looser regulations. Speaking on the sidelines of a Financial Times conference in London, Huang said the U.S. government’s new rules on AI could lead to dozens of new regulations, while China’s subsidies for energy have made it easier for Chinese companies to operate datacenters containing their homegrown AI chips.
Nvidia to Supply Samsung With 50,000 GPUs
Source: The Information Nvidia said Thursday it signed a deal to sell 50,000 graphics processing units to Samsung Electronics to help the South Korean company with its chip design and manufacturing.
Could Apple and Musk’s SpaceX Finally Do a Satellite Deal?
A few years ago, Elon Musk pitched Apple unsuccessfully on the idea of his company, SpaceX, providing satellite connectivity to iPhones. There are tantalizing hints that such a deal could be back on the table.
Apple CEO Wraps Up China Visit as iPhone Air Preorders Start
Source: The Information Apple’s iPhone Air became available for preorders in China Friday, after a delay due to regulatory approvals. The iPhone Air’s China debut, about a month after the phone’s U.S. launch, comes just after CEO Tim Cook visited the country earlier this week. During his China trip, Cook met with senior officials at the Ministry of Industry and Information Technology, which oversees telecom operators, and the Ministry of Commerce, according to the Xinhua news agency.
Anthropic to Open India Office Next Year as Local Demand Grows
Source: The Information Anthropic, the startup known for its Claude AI models, is accelerating international expansion with a plan to open an office in India by next year, the second in Asia after Tokyo. Anthropic CEO Dario Amodei is visiting India this week to meet with public officials and enterprise partners, the company said in a blog post on Wednesday.
OpenAI Launches Cheaper Subscription Tier for Indian Users
China and India should be partners, not adversaries, says foreign minister Wang Yi India and China should view each other as "partners" rather than "adversaries or threats", Chinese Foreign Minister Wang Yi said on Monday, as he arrived for a two-day visit to Delhi. Yi met with Indian Foreign Minister S Jaishankar - only the second such meeting between the two sides since 2020 - …
OpenAI Launches Cheaper Subscription Tier for Indian Users
Source: The Information OpenAI has launched a special subscription tier for Indian users for 399 rupees ($4.60) a month, confirming The Information’s earlier reporting. The new tier, called ChatGPT Go, gives Indian users higher message limits, image generations, uploads and memory, OpenAI Chief Product Officer Kevin Weil said on X on Monday. However, it doesn’t offer as many features as ChatGPT Plus, the more premium version of the chatbot that costs $20 a month.
Chip Startup Rivos Seeks Up to $500 Million to Compete With Nvidia
Rivos, a chip startup that aims to compete with Nvidia, is seeking to raise between $400 million and $500 million, according to two people familiar with the matter. The Santa Clara, Calif.–based startup, backed by Intel’s CEO Lip-Bu Tan, is seeking a valuation of more than $2 billion, one of the people said. It’s not clear if the valuation would include the new investment.
Why Apple Is Losing Ground in the AI Talent War (It’s Not Just Money)
Earlier this year, the Apple team working on the company’s artificial intelligence models wanted to release several of them as open-source software. Doing so would have shown Apple’s technical progress in AI while also harnessing the help of outside researchers to improve the models. Still, it also would have revealed to the public how dramatically the performance of the models dropped when Apple shrunk them to fit on iPhones, compared to versions for more powerful PCs or data center computers.
U.S. Lawmaker Presses Commerce Over Nvidia’s AI Chip Sales to China
Source: The Information The chair of the U.S. House Committee on China has raised concerns over Nvidia’s resumption of sales of its H20 artificial intelligence chip to China. In a letter addressed to Commerce Secretary Howard Lutnick, Michigan Rep. John Mollenaar asked for a briefing from the Commerce Department over the issue.
ByteDance Developing Mixed Reality Goggles in Challenge to Meta
Competition in mixed reality devices is heating up. TikTok’s parent, ByteDance, is working on a lightweight mixed reality device that resembles a pair of goggles, according to three people with direct knowledge of the project. It’s following in the footsteps of Meta Platforms, which has been working on a similar product scheduled for release in 2027. Mixed reality devices let people overlay digital images over their real surroundings to enhance productivity and provide entertainment.
Microsoft Scales Back Ambitions for AI Chips to Overcome Delays
Microsoft is revising its roadmap for its internally developed artificial intelligence server chips and will focus on releasing less ambitious designs through 2028, hoping to overcome problems that have caused delays in development, according to two people with direct knowledge of the situation. Microsoft hopes that by scaling back some of the designs and pushing out the schedule for other AI chips it is already working on, it can develop them more easily.
Microsoft’s AI Chip Effort Falls Behind
Microsoft has spent years designing its own artificial intelligence chips, in part to reduce its dependence on Nvidia. It isn’t going so well. The software giant, which Nvidia employees say was its largest customer by revenue last year, is taking much longer than expected to design its latest-generation AI chips, raising the possibility that the new chips will be even less competitive with Nvidia’s by the time they go into mass production.
Nvidia Working on New Chip for China That Can Form Computing Cluster
Source: The Information Nvidia is developing a new chip for China that for the first time will allow customers to connect multiple chips together to create higher-performance computing clusters, The Information reported. The move comes as the U.S. tightens restrictions on sales of advanced chips to China, though the guidelines have focused on the performance of a single chip. Nvidia has repeatedly complied with the U.S. export controls by redesigning its chips for China.
Nvidia CEO Meets With Trump Regularly to Influence China Chip Policy
Source: The Information Nvidia CEO Jensen Huang has been meeting with President Donald Trump more than what’s publicly known in an effort to persuade the administration to roll back restrictions on its sales to China, The Information reported. The move comes as Trump’s second term has upended the dynamics of Nvidia’s lobbying efforts, requiring Huang to become more engaged with U.S. officials.
Jensen Huang Used to Delegate Politics—Until Trump’s Return
Since the start of this year, Nvidia CEO Jensen Huang has been thrust into an unexpected new role—chief lobbyist for his company—forcing him to carry its hopes, suggestions and pleas directly to President Donald Trump. Many of Huang’s conversations with Trump have happened at Mar-a-Lago, Trump’s residence and private club in south Florida, and Huang has made far more visits there than have been publicly reported, according to a person with direct knowledge of the matter.
Apple’s Satellite Ambitions Threatened by Elon Musk, Internal Resistance
Three years ago, Elon Musk approached Apple with an eleventh-hour offer. Musk had heard that Apple was about to announce a feature for the upcoming iPhone 14 in partnership with satellite firm Globalstar that would allow iPhones to send text messages to emergency services in areas without cellular reception. He wanted Apple to instead use satellite internet service from Starlink, a Globalstar rival operated by Musk’s rocket company, SpaceX.
Show More
loading
Actions
Get in touch with Wayne
Contact Wayne, search articles and posts on X, monitor coverage, and track replies from one place.
Learn more about Muck Rack