YOUR DAILY EDGE: 30 July 2025: AI Corner

June JOLTS: Treading Water

The June JOLTS report offered the latest signs of tepid demand for labor. Job openings declined to 7.4 million in June, coming off the heels of a slight downward revision to May’s unexpected jump. The pullback was presaged by the sharp slowing in private sector hiring in June, which was buried underneath the payroll report’s better-than-expected headline gain in June. The resumption of the slide in job openings better aligns the JOLTS data with separate data from Indeed that indicate job postings continue to gradually slide and small business hiring plans that remain stuck near cycle lows. (…)

Persistently low churn also leaves the labor market looking more fragile than headline numbers suggest. The hiring rate fell to a seven-month low of 3.3% in June. Despite the reluctance to bring on new workers, net employment gains have held up thanks to layoffs remaining low. The layoff rate was unchanged at 1.0%, keeping it well below its pre-pandemic average of 1.2%. (…)

 

• June 2025 JOLTS Report: Running on Tired Legs

Today’s JOLTS report showed that the labor market remained resilient in June, but continued to cool gradually — a sign that fatigue may be building.

Layoffs remained at historically low levels, but job openings were revised down in May and fell further in June. Hiring and quitting activity also remained muted, which hardly reflects an economy that is dynamically adding jobs or instilling confidence in workers.

Right now, the labor market is like a series of runners at mile 20 of a marathon. Some runners are still keeping pace using old momentum, but we’re starting to see many drop off. Information is showing early signs of hiring reacceleration as investment in and adoption of AI technologies grows, while the healthcare industry continues to press ahead at a solid pace.

However, other industries, like wholesale trade and financial activities, have faded over the last year. Overall, the labor market is not collapsing, but it is starting to run on tired legs. In the months ahead, whether this slow fade becomes a stumble will depend on whether demand finds a second wind or if fatigue takes over.

ECB’s Wage Tracker Signals Cooling Pay Pressures Into Next Year

The ECB’s wage tracker, published Wednesday, predicts salaries will rise by an annual 1.7% in the first quarter of 2026. That’s well below a peak of 5.2% recorded at the end of 2024.

The downward trend reflects the “impact of large one-off payments — that were paid in 2024 but drop out in 2025 — and the frontloaded nature of wage increases in some sectors in 2024,” the ECB said in a statement. It sees pay growth averaging 3.2% in 2025. (…)

AI CORNER

How China Is Girding for an AI Battle With the U.S. As Washington tries to limit China’s progress, Beijing is spending more to build AI that doesn’t rely on U.S. technology

(…) Many of the initiatives were on display at an AI conference that ended this week in Shanghai, which Chinese authorities used as a showcase for products free of U.S. technologies.

One startup, Shanghai-based StepFun, touted a new AI model that it said required less computing power and memory than other systems, making it more compatible with Chinese-made semiconductors. Although Chinese chips are less capable than American products, Huawei Technologies and other companies have been narrowing the gap by clustering more chips together, boosting their performance.

China also released an AI global governance plan at the event, the World Artificial Intelligence Conference, which called for establishing an international open-source community through which AI models can be freely deployed and improved by users. Industry participants say it showed China’s ambition to set global standards for AI and could undermine the U.S., whose leading models aren’t open-source.

The conference followed a series of announcements and investments in China aimed at turbocharging its AI capabilities, including rapid expansions in power generation and skills training.

The whole-nation effort, led by Beijing, includes billions of dollars in spending by state-owned enterprises, private companies and local governments. (…)

The rising popularity of DeepSeek, the Chinese AI startup, has buoyed Beijing’s hopes that it can become more self-sufficient. Huawei has published several papers this year detailing how its researchers used its homegrown chips to build large language models without relying on American technology. (…)

China’s biggest AI challenge is overcoming its difficulty in sourcing the world’s most-advanced chips. Washington has cut China off from some of Nvidia’s most sophisticated semiconductors, as well as the advanced machine tools used to make cutting-edge chips, restrictions that many experts believe will continue to hold China back.

Huawei is helping spearhead efforts to navigate those restrictions. During a meeting with President Xi Jinping in February, Chief Executive Officer Ren Zhengfei told Xi about “Project Spare Tire,” an effort by Huawei and 2,000 other enterprises to help China’s semiconductor sector achieve a self-sufficiency rate of 70% by 2028, according to people familiar with the meeting.

Increasingly, the company has been able to bundle together the best chips it can produce to match the performance of some American computing systems. That is helping local companies reach some of the same computing goals as the U.S., such as training state-of-the-art generative AI models, though it consumes more power than U.S. chips.

U.S. researcher SemiAnalysis recently reported that one such Huawei cluster, which connects 384 of its Ascend chips, outperforms Nvidia’s flagship system with 72 graphics-processing units on some metrics.

Morgan Stanley analysts forecast that China will have 82% of AI chips from domestic makers by 2027, up from 34% in 2024. (…)

Prodded by Beijing, Chinese financial institutions, state-owned companies and government agencies have rushed to deploy Chinese-made AI models, including DeepSeek and Alibaba’s Qwen. That has fueled demand for homegrown AI technologies and fostered domestic supply chains.

Some companies ordered expensive AI servers to deploy the models even before they came up with a use for the technology, according to people familiar with the matter.

At the Shanghai conference, organized by China’s government and featuring more than 800 companies, Chinese researchers played down the impact of U.S. export controls on advanced chips. They compared notes on how they were focusing on overcoming bottlenecks through better model designs and engineering techniques.

Many also touted Chinese companies’ tendency to give users free access to modify and deploy their AI models, an open-source approach that has boosted adoption of Chinese models globally. 

While the world’s best large language model is still American, the best model that everyone can use free is now Chinese. According to benchmark provider Artificial Analysis, the overall performance of China’s best open-weight model has surpassed the American champion since November.

In recent weeks, a flurry of Chinese companies have flooded the market with open-source models, many of which are claiming to surpass DeepSeek’s performance in certain use cases. OpenAI’s Sam Altman said his company had pushed back the release of its open-source model indefinitely for further safety testing.

China is also investing heavily in other areas, including more electricity to power domestic data centers to develop and run AI.

It is spending $564 billion on grid construction projects in the five years up to 2030, an increase of more than 40% from the previous five years, Morgan Stanley researchers forecast. 

China currently has about 2.5 times as much power-generation capacity as the U.S., a disparity that is projected to grow larger in the next five years despite the U.S. expanding power generation.

China has also approved more than 600 colleges to set up degree programs in AI, according to data released by the country’s Ministry of Education in April. That is up from 35 universities with such programs in 2019.

In Beijing, primary and secondary schools will begin mandatory AI lessons for students starting in September.

China’s efforts have already enabled it to develop a robust pipeline of homegrown talent, according to researchers from the Hoover Institution and Stanford University, who recently evaluated the backgrounds of more than 200 authors involved in DeepSeek’s papers between 2024 and February 2025. More than half of these DeepSeek researchers never left China for schooling or work, they found.

The U.S. has fewer universities that offer AI degree programs, but American universities dominate rankings for computer and information science. This April, President Trump signed an executive order mandating AI education and learning opportunities for American youth.

Yesterday, Reuters informed us that

(…) The “Model-Chip Ecosystem Innovation Alliance” brings together Chinese developers of large language models (LLMs) and AI chip manufacturers.

“This is an innovative ecosystem that connects the complete technology chain from chips to models to infrastructure,” said Zhao Lidong, CEO of Enflame, one of the participating chipmakers.

Other manufacturers of graphics processing units (GPUs) in the alliance include Huawei, Biren, and Moore Threads, which have been hit by U.S. sanctions that block them from purchasing advanced tech made with U.S. know-how. The alliance was announced by StepFun, an LLM developer.

A second alliance, the Shanghai General Chamber of Commerce AI Committee, aims to “promote the deep integration of AI technology and industrial transformation.” (…)

One of the most talked about products at the conference was Huawei’s CloudMatrix 384 which incorporates 384 of its latest 910C chips and outperforms Nvidia’s GB200 NVL72 on some metrics, according to U.S. research firm SemiAnalysis.

Huawei’s system design capabilities have meant that it has been able to use more chips and system-level innovations to compensate for weaker individual chip performance, SemiAnalysis said.

At least six other Chinese computing firms showcased similar “clustering” chip technology. Metax demonstrated an AI supernode featuring 128 C550 chips designed to support large-scale liquid-cooled data centre requirements. (…)

• China’s AI Strategy Relies on Frenzy and Frenemies

Held on the banks of the Huangpu river in Shanghai, the World AI Conference convened thousands of people — as well as scores of robots — and brought to life all the passions and pitfalls of the current state of AI in China. It also put into stark contrast the chasm between the strategy pushed by Beijing and the one touted by the White House.

It’s the first major gathering since DeepSeek’s breakthrough reasoning model launched earlier this year, driving intense competition at home and proving China can go toe-to-toe with Silicon Valley. With that exuberance came the crowds of challengers, present in so many domestic industries, encouraged by government support and an open-source ecosystem that allows firms to quickly learn from rivals. When one of the so-called Little Dragons, Moonshot, released a massive open-source model that excelled at coding tasks, Alibaba Group Holding Ltd. was able to update their own Qwen model within about a week to improve benchmarks at the same skills that sent Kimi-K2 viral.

Beijing likes to say this approach democratizes access to AI by offering the world the ability to freely build atop its tools, and it gives local developers an edge. (…)

China’s big gathering kicked off just days after President Donald Trump pledged that that the US will “do whatever it takes to lead the world in artificial intelligence.” After unveiling a so-called AI Action Plan, he went on to declare that America is the country that started the AI race, and “is going to win it.”

In Shanghai, Premier Li Qiang headlined opening night by announcing that China will organize the launch of an international body to jointly develop the technology, with the goal of preventing it from becoming “an exclusive game for a small number of countries and enterprises.” It dovetailed with this year’s conference theme: “Global solidarity in the AI era.” China is willing to share its development and products with the world, Li said, especially in the Global South. (…)

Beijing’s approach relies on convincing the world to use its plethora of low-cost AI products being rapidly released. Fresh access to Nvidia processers is giving the industry new momentum, and the increasingly crowded field is driving down prices. Compare the two and China’s plan seems more strategic in the long run. (…)

The FT’s Julian Gewirtz yesterday contrasted the strategies:

(…) Both countries are investing heavily in the technology. But, outwardly at least, they appear to be pursuing different goals. While US leaders prioritise developing the most intelligent models possible, Chinese policymakers are most focused on AI’s widespread application. To use Silicon Valley parlance, the Chinese Communist party seems far less “AGI-pilled” than their American counterparts. Although leading Chinese labs like DeepSeek, Zhipu, and Stepfun state their ambitions to reach AI that can equal the abilities of the human mind, top Chinese government officials barely address AGI.

For one, the Chinese government lacks an official term that captures what Altman and others mean when they say AGI. The term used in China is tongyong rengong zhineng, the precise translation of which is “general-purpose AI”. This suggests a system applied to many uses rather than human-level intelligence.

Nor is the Chinese government taking obvious steps to create the physical infrastructure necessary for AGI. It has been slow to build the “superclusters” of computing chips that US hyperscalers are constructing.

Instead, Xi Jinping has emphasised applying AI to practical purposes. He has consistently said that China’s AI sector should be “strongly oriented toward applications.” From industrial robotics to products like Manus, a leading AI agent, this is already becoming a reality. (…)

Also: “Last Friday, Alibaba released the latest version of its Qwen open source LLM models, the Qwen3-Thinking-2507.  According to VentureBeat, the new Qwen “now leads or closely trails top-performing models across several major benchmarks.” (Fortune)

FYI:

• AI is already touching nearly every corner of the medical field 
• AI isn’t just entering law offices—it’s challenging the entire legal playbook 
• How a bulldozer, crane, and excavator rental company is using AI to save 3,000 hours per week 
• Meet the companies using AI to reinvent the energy business 
• How Walmart, Amazon, and other retail giants are using AI to reinvent the supply chain—from warehouse to checkout

AI Search Is Growing More Quickly Than Expected Large language models aren’t replacing traditional browsers anytime soon, but they have become another responsibility for brands

Chatbots are becoming the go-to source for online answers for many consumers, chipping away at the dominance of traditional web search and adding another avenue of outreach that brands must cultivate to connect with customers.

An estimated 5.6% of U.S. search traffic on desktop browsers last month went to an AI-powered large language model like ChatGPT or Perplexity, according to Datos, a market intelligence firm that tracks web users’ behavior.

That pales beside the 94.4% that still went to traditional search engines like Alphabet’s Google or Microsoft’s Bing, which have tried to fight off the new competition by adding artificial intelligence summaries to the top of their search results.

But the percentage of traffic that went to browser-based AI search has more than doubled since June 2024, when it was 2.48%, according to Datos, which is part of marketing software company Semrush. It has more than quadrupled since January 2024, when the figure was just under 1.3%. (…)

The numbers exclude activity on mobile browsers and apps including those from OpenAI’s ChatGPT and Google.

The rapid growth in AI searches could mark a sea change in online behavior comparable to the emergence of Google’s web browser and the first social-media platforms, according to Eli Goodman, chief executive and co-founder of Datos.

The numbers are more striking among so-called early adopters, or consumers who had already started using LLMs in desktop browsers when Datos began tracking their behavior in April 2024.

Forty percent of desktop browser visits among these early adopters went to LLMs, up from just over 24% in June 2024, Datos research found. Traditional search engines’ share of the desktop browser traffic from these early adopters fell significantly in the same period, to 61% in June from around 76% a year earlier.

The amount of time that consumers worldwide spent on traditional search apps and websites declined 3% from April 2024 to April 2025, according to a report released last month by market research firm Sensor Tower. The drop was twice as high among early adopters, defined as people who first downloaded ChatGPT in 2023, the report said.

Use of Google’s traditional search product is still growing, and the AI overviews that now appear on Google searches are leading to more queries that connect consumers to businesses, a Google spokeswoman said.

The growth in AI searches, plus the appearance of AI overviews in traditional search browsers, is likely to further depress search traffic to the websites where brands have long directed energy, resources and search advertising, according to Neil Vogel, CEO of Dotdash Meredith, the publisher of magazine and digital brands including People and Better Homes & Gardens. (…)

For brands, the most important difference between LLMs and browsers is that LLMs reveal one answer rather than a list of links, giving businesses fewer openings to appear before consumers, said Vogel. (…)

Search engines still handle an overwhelming majority of overall search traffic and remain firmly embedded in the smartphones that occupy so much of our time, he said.

Brands must also remember that AI searches often serve different needs than a traditional google search, he said.

“Over 90% of all of the AI searches are what we call informational or productivity-based: Help me solve this problem, help me answer this question,” said Goodman. Traditional search result pages, on the other hand, were designed to lead consumers out to other destinations online, he said. (…)

OpenAI says it has no plans to develop ad products. Perplexity has begun experimenting with sponsored searches, however, and last year hired a head of advertising and shopping.

Perplexity this year also released its own web browser, where searches are handled by its AI instead of a traditional engine like Google or Bing. The company says its search gives users “the choice to navigate the web.”

OpenAI is close to releasing a browser as well, Reuters reported this month. The company declined to comment.

On AI, just about everything will grow faster than expected, including China…

The US is particularly vulnerable on energy, AI’s essential and very hungry backbone. I wrote on that on September 23, 2024 (Power Play). Things are actually getting worse for the US.

• Electricity consumption from data centres, artificial intelligence (AI) and the cryptocurrency sector could double by 2026. Data centres are significant drivers of growth in electricity demand in many regions. After globally consuming an estimated 460 terawatt-hours (TWh) in 2022, data centres’ total electricity consumption could reach more than 1 000 TWh in 2026. This demand is roughly equivalent to the electricity consumption of Japan. (IEA)
• China generated over 10,000 TWh (terawatt-hour) of electricity in 2024. That’s more than the combined output of the U.S., EU, and India—the next three biggest producers.
• China’s capacity, already more than 2x the US, is growing very rapidly and at lower costs.

• Renewables are set to provide more than one-third of total electricity generation globally by early 2025, overtaking coal. The share of renewables in electricity generation is forecast to rise from 30% in 2023 to 37% in 2026, with the growth largely supported by the expansion of ever cheaper solar PV. Through this period, renewables are set to more than offset demand growth in advanced economies such as the United States and the European Union, displacing fossil-fired supply. At the same time, in China, the rapid expansion of renewable energy sources is expected to meet all additional electricity demand. (IEA)
• America was already losing to China on clean energy. Trump just sealed its fate

The new clean energy regime can be summarized in one incredible statistic: China installed more wind and solar power in a single year than the total amount of renewable energy currently operating in the United States.

America was already laps behind China in the race to dominate the industry, new data from Global Energy Monitor shows. President Donald Trump’s “big, beautiful,” spending bill will secure its position as a clean-energy loser, experts told CNN.

The spending law Trump signed earlier this month knee-caps clean energy tax credits for wind and solar. Business leaders say it will raise electricity prices for businesses and consumers alike here, as the cheapest electrons on the grid (generated by wind and solar) become more costly to build and are replaced with more expensive gas.

At the same time, pulling funds from the clean energy industry puts it on its heels just as it was looking to make gains toward more efficient technologies and better battery storage.

Meanwhile, China is currently building 510 gigawatts of utility-scale solar and wind capacity, according to Global Energy Monitor. It will be added to the eye-popping 1,400 gigawatts already online — five times what is operating in the US.

Wind and solar, bolstered by giant batteries that can store their energy, are also becoming an increasingly dominant force in the US, but on a much smaller scale. Renewables generate the vast majority of new electricity that’s come online in the past few years in the US and make up about 85% of what is currently waiting to be approved in the nation’s permitting queue. (…)

The law effectively cuts planned renewables additions to the grid in half over the next decade compared to projections without it, according to modeling done by the non-partisan think tank Rhodium Group. That will mean rising electricity prices in every continental US state, due to the price of renewables increasing and more expensive gas filling the gap, as CNN has reported. (…)

Meanwhile, in the US, more expensive electricity could significantly hamper economic development and discourage companies from building here, undermining one of Trump’s own priorities. With less wind and solar coming online due to the GOP law, plus long wait times to get new natural gas plants up and running, Rhodium analyst Ben King said some data centers and large manufacturing facilities may struggle to get enough power.

“Data centers, semiconductor manufacturing and other sources of new industrial load, that just might not be able to come online, because we may not have the generators to meet that demand,” King said.

BTW, about the other AI backbone:

• Trump signs order gutting Department of Education
• Harvard has expressed a willingness to spend up to $500 million to settle its dispute with the White House, which has accused the university of civil rights violations tied to antisemitism and DEI policies, The New York Times reports.
• Trump officials see Columbia’s settlement as a playbook for negotiations with other universities, combining financial penalties with internal policy changes and external oversight. Emboldened by its early success, the Trump administration this week launched new investigations and lawsuits targeting UCLA, Duke, and George Mason. (Axios)

Adidas warns of €200mn tariff hit Shares in German sportswear maker slide as Donald Trump’s tariffs add to costs

Adidas has warned that US tariffs will increase its costs by up to €200mn in the second half of the year (…). Chief executive Bjørn Gulden said the tariffs had already cost the company “double-digit euro millions” in the second quarter (…).

  FYI: According to Statistics Canada, the number of Canadians crossing the border by plane fell through the first five months of the year, and was down in April and May by 14 per cent and 24 per cent, respectively, compared to a year ago. Trips by car, which comprise three-quarters of travel to the U.S., fell by more than 35 per cent for these months.