Source: Our World in Data

Made with 🩵 by Dialid Santiago.

Just read this fascinating article in The New Yorker

What Happens After A.I. Destroys College Writing?

The demise of the English paper will end a long intellectual tradition, but it’s also an opportunity to reëxamine the purpose of higher education.

https://www.newyorker.com/magazine/2025/07/07/the-end-of-the-english-paper

Takeaways:

🔄 A.I. shifts our opportunity costs. “College is all about opportunity costs.” Students today spend far less time on schoolwork than in decades past. In the early 60s, college students spent an estimated 24 hours a week on schoolwork. Today, that figure is about 15 hours. With tools like ChatGPT, what used to take hours now takes minutes. But what are we giving up when we outsource the messy, formative process of thinking?

⚡ The intoxication of hyperefficiency. Most students start using A.I. as an organizer aid but quickly evolve to off-loading their thinking altogether. Moreover, they describe using it like social media: constantly open, constantly tempting. Are we becoming so efficient that we forget why we’re thinking in the first place?

✍️ Yes, typing is fast… “but neuroscientists have found that the “embodied experience” of writing by hand taps into parts of the brain that typing does not. Being able to write one way—even if it’s more efficient—doesn’t make the other way obsolete.”

🧍‍♂️ What does it mean to “sound like ourselves”? As AI gets better at sounding like us, do we risk forgetting what we sound like in the first place?

📉 Is Cognitive decline associated to A.I.? According to a recent study from the Organisation for Economic Co-operation and Development, human intellect has declined. The assessment of tens of thousands of adults across 31 countries showed an over-all decade-long drop in test scores for math and for reading comprehension. Andreas Schleicher, the director for education and skills at the O.E.C.D., hypothesized that the way we consume information today—often through short social-media posts—has something to do with the decline in literacy.

Just finished watching Geoffrey Hinton’s interview on The Diary of a CEO — a rare, candid conversation from one of the pioneers of AI.

Takeaways

1. Real AI risks are already here Hinton emphasizes that the pressing dangers aren’t sci-fi superintelligence, but things we see today: algorithmic manipulation, the amplification of bias, and disinformation at scale.

2. Hinton comes from a family tree of scientific legacy. His lineage includes George Boole best known as the author of The Laws of Thought (1854), which contains Boolean algebra; George Everest, a geographer and a Surveyor General of India, after whom Mount Everest was named; and Joan Hinton a nuclear physicist and one of the few women scientists who worked for the Manhattan Project in Los Alamos.

3. Hinton left his academic post not for ambition or prestige, but to earn enough to support his son, who has learning disabilities. His story highlights a quiet crisis: academia struggles to retain top talent when it can’t offer the financial security that industry can. What does it say about our priorities when some of the most important work in science can’t afford to support a family?

Related

Why Geoffrey Hinton is sounding the alarm about AI?, Toronto Life

Rage Against The Machine: Geoffrey Hinton spent half a century developing artificial intelligence. Now, he worries that his life’s work could spell the end of humanity. Inside his mission to warn the world.

https://torontolife.com/deep-dives/geoffrey-hinton-sounding-alarm-artificial-intelligence/

Cartea, Álvaro and Jin, Qi and Shi, Yuantao, The Limited Virtue of Complexity in a Noisy World (April 02, 2025). Available at SSRN here or here.

Summary: In this paper, the authors analyse the role of model complexity in the context of predicting asset returns and portfolio construction. In particular, they aim to address the significant question of whether adding a large number of predictive features ultimately harms performance. Their work aims to bridge two views: the traditional econometric one, favoring parsimonious models (i.e. Occam’s razor), and the more modern machine learning findings that highly the fact that so-called “overparameterized” models can perform well under proper regularization ( double descent phenomenon).

The authors set up a framework where investors predicts excess returns using a large number of features, but these features are “contaminated” by noise (which can arise from data collection gaps, computational approximations, or other infrastructure limitations). They examine how this affects the Sharpe ratio of a timing strategy, which the investor seeks to maximize, and the out-of-sample R-squared of return forecasts. Given the high-dimensional setting, they use ridge regression and apply Random Matrix Theory classical results to characterise the asymptotic behaviour of these metrics, focusing on the case where true features are independent (a quantitative investing).

Their results show that model complexity can improve asset return predictions and portfolio performance when regularisation is used and data quality is high. However, when features are noisy or only partially observed, there exists an optimal level of complexity. Beyond this point, adding features introduces more noise than signal, degrading both predictive accuracy and portfolio outcomes.

Takeaway: In asset return prediction, higher complexity is not always better. When data are noisy, more features can harm performance underscoring that garbage in, garbage out applies strongly in quantitative finance.

Natural History Ensemble, no. 11 (1596–1610) by the talented Flemish draughtsman Anselmus Boëtius de Boodt (1550–1632). Check out more of Anselmus Boëtius de Boodt’s works in this Rawpixel Gallery.

Source: Public Domain. Original from the Rijksmuseum.