Remarkable 4

Some interesting stuff we found this week, although normally a weekend read, so a little late.

Nobel prize-winning economist Daren Acemoglu on AI:

Acemoglu said he sees great potential in the technology. But he believes its current trajectory won’t deliver that promise, in large part because it’s being applied to the wrong types of problems in the wrong professions. The focus, he suggests, should be on the provision of truly reliable information within specific problem-solving contexts. He points to a wide range of roles in which AI use is largely absent, including electrician, plumber, nurse, educator, and clerical worker. People in such professions are typically “engaged in problem-solving tasks,” he writes. “These tasks require real-time, context-dependent and reliable information.” Think of an electrician who has to deal with an equipment malfunction or a short-circuit on the electricity grid but lacks sufficient knowledge to troubleshoot the problem. If AI tools can provide this kind of information when such workers need it — something the tools aren’t currently reliable enough to do (think hallucinations) — then the technology’s economic impact could be significantly greater, Acemoglu writes.

A new look at the economics of AI

Another Nobel Prize-winning economist, Paul Krugman, interviews Paul Kedrosky, investor, tech expert, and research fellow at MIT, who has some worrying things to say about the AI-CapEx build-out (some along the lines of our own hesitations). The whole interview is worth reading until the end, and it’s an uncomfortable read.

Kedrosky: We have this prodigious amount of spending going on, and that was one of the windows through which I got interested in the investment side of this stuff, because it seemed as if it was so large that it was having an impact on economic data itself. I was looking at that early this year—I just did yesterday or the day before—there was a new OECD report on the US showing that in the first half of 2025, the US was arguably in a recession absent AI CapEx spending which was scarcely a ripple in terms of people saying, “hello, we’re running a giant private sector stimulus program that’s keeping the US out of recession,” and yet no one’s talking about it in those terms.

Talking With Paul Kedrosky - Paul Krugman

We have brain circuitry for motivation:

We all know people with very different levels of motivation. Some will go the extra mile in any endeavour. Others just can’t be bothered to put the effort in. We might think of them as lazy – happiest on the sofa, rather than planning their latest project. What’s behind this variation? Most of us would probably attribute it to a mixture of temperament, circumstances, upbringing or even values. But research in neuroscience and in patients with brain disorders is challenging these assumptions by revealing the brain mechanisms that underlie motivation. When these systems become dysfunctional, people who were once highly motivated can become pathologically apathetic. Whereas previously they might have been curious, highly engaged and productive – at work, in their social lives and in their creative thinking – they can suddenly seem like the opposite.

Does ‘laziness’ start in the brain?

The IMF worries about shadow banks:

The head of the International Monetary Fund has admitted that worrying about the risks building up in non-bank lending markets keeps her awake at night. Kristalina Georgieva on Thursday urged countries to pay more attention to the private credit market, after the failure of the sub-prime auto lender Tricolor and the car parts supplier First Brands. Speaking at the IMF’s annual meeting in Washington DC, Georgieva said the fund was concerned about the “very significant shift of financing” from the banking sector to non-bank financial institutions (NBFIs). Those NBFIs are not regulated as closely as the banking sector, she pointed out, meaning the world could end up in “a difficult place” if the private credit sector continued to grow significantly and the global economy then weakened.

Head of IMF says risks in private credit market keep her awake at night

The effects of AI on the job market:

As many as three million jobs could be lost over the next decade due to the rapidly increasing role played by AI and automation, a major educational research organisation has warned. A report by the National Foundation for Educational Research (NFER) warns that “extensive changes are required” to ensure people have the skills needed to maintain a foothold in the workplace. The areas particularly at risk are in “high-risk declining occupations”, the research team said, suggesting administrative, secretarial, customer service and machine operators may be among the most threatened roles.

AI could replace 3 million jobs over next decade, report warns

We weren’t aware of this debate:

The debate over whether or not to wash clothes before wearing them has raged among shoppers for years. But experts seem to agree, for the most part, on washing beforehand. Frances Kozen, senior lecturer in fiber science at Cornell University told Real Simple she typically washes before wearing “to remove processing chemicals, excess dye, and dirt from all of the handling during production.” “Textiles and garments have passed through many hands, often in multiple countries,” Kozen added.

Experts clarify the clothes you should probably wash before wearing | The Independent

How to open the black box that is AI at least a little:

As AI systems enter production, reliability and governance can’t depend on wishful thinking. Here’s how observability turns large language models (LLMs) into auditable, trustworthy enterprise systems. Why observability secures the future of enterprise AI The enterprise race to deploy LLM systems mirrors the early days of cloud adoption. Executives love the promise; compliance demands accountability; engineers just want a paved road. Yet, beneath the excitement, most leaders admit they can’t trace how AI decisions are made, whether they helped the business, or if they broke any rule. Take one Fortune 100 bank that deployed an LLM to classify loan applications. Benchmark accuracy looked stellar. Yet, 6 months later, auditors found that 18% of critical cases were misrouted, without a single alert or trace. The root cause wasn’t bias or bad data. It was invisible. No observability, no accountability. If you can’t observe it, you can’t trust it. And unobserved AI will fail in silence. Visibility isn’t a luxury; it’s the foundation of trust. Without it, AI becomes ungovernable.

Turning AI from experimental to operational starts with true observability.

Keeping an eye on this one:

The chip war situation has developed not necessarily to America’s advantage. China’s 14th Five-Year Plan $15B quantum push has produced millions of photonic quantum chips that are now solving problems in hospitals and laboratories, but using 10% of the power used by an electronic chip and running 1000x times faster. For the first time, optical quantum computers are industrial-grade products.

China’s Photonic Chips End America’s IC Dominance. Permanently?

Count us in the HSBC camp on this:

Per the Financial Times (LINK), HSBC has serious doubts about OpenAI’s financial wherewithal. The following bullet points outline HSBC’s assumptions, which highlight the challenging financial path OpenAI faces.

• OpenAI recently committed to renting up to four additional gigawatts of cloud computing space from Microsoft and Amazon, bringing its total contracted power to 36GW

• Based on all its known data center rental deals, including recent agreements with Microsoft, Oracle, CoreWeave, and Amazon, OpenAI’s rental expenses will total up to $1.8 trillion. HSBC forecasts that this will translate into roughly $620 billion in annual rental costs upon full deployment.

• Even with extremely bullish usage assumptions including (3 billion users or 44% of the worlds adult population by 2030, 10% of users having paid subscriptions, revenues from search and advertising revenue, and strong enterprise growth) OpenAI’s projected cumulative free cash flow through 2030 reaches only about $282bn, far short of the cumulative $792bn in rental costs HSBC expects over the same period

• After adding Nvidia’s promised investments, existing cash, and undrawn financing facilities, HSBC still sees a $207 billion funding gap through 2030, meaning OpenAI will need repeated significant capital raises to keep operating its rented data centers

Daily Market Trading Update: December 1, 2025

Chinese startup DeepSeek has released the first ever open AI model capable of achieving a gold medal at the prestigious International Mathematical Olympiad. The Math-V2 model reached the gold-level score at this year’s IMO – a feat that only 8 per cent of human participants achieved – by demonstrating its reasoning abilities rather than just produce a simple answer. DeepSeek has now made the model available to the public on the developer platforms Hugging Face and GitHub, allowing anyone to run or modify it for free.

China’s DeepSeek releases first AI capable of top score at maths olympiad | The Independent

Humanity will have to decide by 2030 whether to take the “ultimate risk” of letting artificial intelligence systems train themselves to become more powerful, one of the world’s leading AI scientists has said. Jared Kaplan, the chief scientist and co-owner of the $180bn (£135bn) US startup Anthropic, said a choice was looming about how much autonomy the systems should be given to evolve. The move could trigger a beneficial “intelligence explosion” – or be the moment humans end up losing control.

‘The biggest decision yet’: Jared Kaplan on allowing AI to train itself

Texas is renowned for its ranching industry, but this oldest of American vocations is fast becoming entwined with icons of modern life: solar and wind farms. Among the cattle, Texas is also home to more than 15,300 wind turbines across 239 projects and 197 utility-scale solar farms. “Solar and wind energy now frequently provide more than 45% of the state’s electricity needs,” says Dennis Wamsted of the Institute for Energy Economics and Financial Analysis.

Gas vs renewables: how will data centre power demand be met?

A single person claims to have authored 113 academic papers on artificial intelligence this year, 89 of which will be presented this week at one of the world’s leading conference on AI and machine learning, which has raised questions among computer scientists about the state of AI research. The author, Kevin Zhu, recently finished a bachelor’s degree in computer science at the University of California, Berkeley, and now runs Algoverse, an AI research and mentoring company for high schoolers – many of whom are his co-authors on the papers. Zhu himself graduated from high school in 2018.

Artificial intelligence research has a slop problem, academics say: ‘It’s a mess’