WSJ: “Startups from Silicon Valley to Europe and beyond are racing to develop cheap, reliable systems to counter hostile drones appearing over airports and global shipping lanes far from the battlefield in Ukraine. The systems explode, collide, fire pellets, jam radios, beam lasers and even shoot nets like Spider-Man. The different approaches reflect the multifaceted threat. Hostile drones come in all shapes, sizes, speeds and altitudes. Some attack while others snoop. They fly alone or in swarms. The setting also dictates the response: Thwarting battlefield attacks in Ukraine requires a different approach from protecting infrastructure against spying drones in a crowded European city. Whatever the situation, equipment must be deployable quickly against fleeting incursions and cost no more than its target.”
Pratap Bhanu Mehta: “[Joel] Mokyr forces us to confront a profound truth: Knowledge does not grow on its own. It is cultivated by cultures and institutions and legitimised by ideas. If India is serious about becoming a knowledge power, it must confront the structural disconnects in its own knowledge systems — the weak bridges between science and engineering, discovery and application, the factory and lab. It must learn to reward and match talent. The real challenge is not just investment or incentives, which we are throwing around freely. We will get easily dazzled by an odd win, ISRO or a new AI centre. The real challenge is building a sustained culture of innovation. Athena, or in the Indian context Saraswati, are not just goddesses to be worshipped: They have to become palpable social forces.”
FT: “One odd thing about AI equipment is that it’s very expensive to buy and very cheap to rent. Want an Nvidia B200 GPU accelerator? Buying one on its release in late 2024 would’ve probably cost around $50,000, which is before all the costs associated with plugging it in and switching it on. Yet by early 2025, the same hardware could be rented for around $3.20 an hour. By last month, the B200’s floor price had fallen to $2.80 per hour. Nvidia upgrades its chip architecture every other year, so there’s an opportunity for the best-funded data centre operators to lock in customers with knockdown prices on anything that’s not cutting edge. From the outside, the steady decline in GPU rental rates resembles the kind of predatory price war the tech industry relies upon: burn money until all your competitors are dead. The evidence, however, is more complicated.”
PwC: “As factories become smarter and more automated, the demand for chips in industrial applications is growing. According to PwC’s “Global semiconductor industry outlook 2026” report, expanded automation in manufacturing will lead to a tenfold increase in the value of chips in a factory. Those chips will perform different functions. At smart factories operating today—with baseline levels of automation—sensors gather data, and programmable logic controllers (PLCs) use chips to remotely control factory equipment. At a smaller scale are microcontroller units (MCUs), which are more customizable and handle precise tasks. At increased levels of autonomy, connectivity integrated circuits (ICs) link IOT equipment throughout the factory for real-time tracking of raw materials to inventory, handling massive amounts of data. Smarter sensors can take over quality control functions, helping to detect defects and adjust machinery. And power-management integrated circuits help keep these systems efficient. Ultimately, automation requires more than just chips on the factory floor. It’s also essential to have more central computing power to make sense of all the data, along with more chips in logistics functions to coordinate inputs and outputs—all powered by AI.”
