This week’s stories are about tech, AI and sustainability, coming from France, Pakistan and the USA

1-Apple loses ground in China’s smartphone rebound

Apple was the only major handset maker to post a year‑on‑year sales slide in China’s resurgent smartphone market last quarter, according to research firm IDC.

• Shipments: Apple moved 9.8 million iPhones in January‑March, down 9 % from the same period in 2024.
• Market share: The U.S. giant’s slice of China’s market slipped to 13.7 %, from 17.4 % in Q4. Apple remains the No. 5 vendor.

I read this story at Reuters and by contrast, Xiaomi—now China’s top seller—shipped 13.3 million handsets, a 40 % jump that helped lift overall industry shipments 3.3 %.

IDC analyst Will Wong attributes Apple’s stumble to its premium pricing. A government subsidy introduced in January rebates 15 % on smartphones priced below 6,000 yuan (≈ US $820), a threshold that excludes most iPhone models but covers many domestic brands, the story noted. (By the way, I have a story here about Apple Turkey)

With lower‑priced competitors benefiting from the incentive, Apple enters Q2 under pressure to defend its high‑end positioning in the world’s largest smartphone market, the story claimed.

.  Apple loses ground in China’s smartphone rebound (Photo: Reuters)

2-AI boom to more than double data‑centre power use by 2030, IEA warns

Artificial intelligence is poised to reshape global electricity demand, with data‑centre power consumption projected to soar past today’s national totals, the International Energy Agency (IEA) says in its new Energy & AI special report.

The IEA forecasts that data‑centre electricity use will jump from about 460 TWh in 2024 to roughly 945 TWh in 2030, more than the annual demand of Japan, the world’s third‑largest economy.

“AI is now one of the biggest stories in energy, yet policymakers have lacked the numbers to grasp its full impact,” IEA Executive Director Fatih Birol said.

Data centres already account for 1.5 % of global power use, and the United States generates 45 % of that load. By 2030, the IEA expects U.S. data centres to drive nearly half of the country’s net growth in electricity demand, eclipsing all other heavy industries combined.

While AI fuels consumption, it can also unlock large efficiency gains. The report estimates that AI‑driven grid optimisation could free up 175 GW of transmission capacity—without stringing new lines—and that AI‑enabled industrial process control could save more energy each year than Mexico currently uses. Meanwhile, AI‑assisted cyber‑attacks on utilities have tripled in four years, the IEA notes.

The agency urges governments and industry to accelerate grid investment, deploy data centres more strategically, and deepen coordination between the energy and tech sectors. To aid decision‑makers, the IEA is launching an interactive “AI Agent” that lets users explore the report’s data and scenarios.

AI boom to more than double data‑centre power use by 2030

3-AI‑powered cameras offer new hope for Pakistan’s endangered snow leopards

A pilot project in Pakistan’s high‑mountain north is testing whether artificial‑intelligence text alerts can cut deadly clashes between villagers and one of the world’s most elusive big cats: the snow leopard.

The World Wide Fund for Nature (WWF) and Lahore University of Management Sciences (LUMS) have deployed ten solar‑powered “smart” camera towers across three villages at roughly 3,000 metres altitude. I saw this story at BBC and trained over three years, the AI model can distinguish snow leopards from humans and livestock. When the system spots a cat, it automatically texts nearby herders, urging them to move sheep and goats under cover.

Globally just 4,000‑6,000 snow leopards remain, and Pakistan hosts about 300— the world’s third‑largest sub‑population, according to the story. WWF estimates 221‑450 cats are killed annually, more than half in retaliation for livestock losses. The resulting 20 % population drop over two decades has pushed the species towards extinction.

Conservationist Asif Iqbal says the cameras have already filmed multiple cats, including a mother marking her territory. Yet accuracy is not perfect: the AI sometimes double‑labels people as “animals,” and fragile mountain cell service can delay alerts, the story emphasized.

Technical fixes—winter‑proof batteries, non‑reflective paint, landslide‑resistant solar panels—took months of trial and error,as the  the story put it. Local scepticism proved tougher. Some residents cut wires or draped blankets over cameras, worried about privacy or doubtful texts would arrive where phones often show “no signal.”

AI‑powered cameras offer new hope for Pakistan’s endangered snow leopards (Photo: BBC)