“We are the unpaid IT directors of our own personal infrastructure. We manage updates, we manage subscriptions, we manage batteries, we manage notifications. We manage the technical debt of our own lives.”

Terry Godier’s reflection on our changing relationship with objects hits a nerve. He uses the classic Casio F-91W as a symbol of the “finished” object—something you buy once, wear for years, and never have to think about again. In contrast, almost every modern gadget we bring home today is a demanding “relationship” that requires constant attention.

The idea that we’ve all become accidental IT managers for the things we own is something we rarely consider when we’re clicking ‘Buy Now.’ We look at the features, but we ignore the invisible job description that comes in the box. We’ve traded the simplicity of ownership for the cognitive cost of keeping our devices “alive”.

Andrej Karpathy was Tesla’s AI director, co-founded OpenAI, and is one of the most respected voices in the field. When he publishes his personal retrospective of the year in LLMs, it’s worth paying attention.

His conclusion is what sticks with me most: LLMs are simultaneously much smarter and much dumber than he expected - and the industry hasn’t yet realized even 10% of their current potential.

In this conversation with Dwarkesh Patel, Ilya Sutskever argues we’re transitioning from the “age of scaling” back to an “age of research.” The logic is straightforward: pre-training is running out of data, and simply throwing more compute at the problem won’t get us to superintelligence — that, according to his estimate, is five to twenty years away, if we solve something more fundamental than scaling.

Kevin Kelly has a provocative thesis: in the near future, authors will pay AI companies to ensure their books are used in training models. Why? Because “if your work is not known and appreciated by the AIs, it will be essentially unknown.”

It’s a strong, almost dystopian claim, but Kelly argues we’re already on that path. He uses his own experience as evidence: he stopped questioning calculators long ago, then stopped questioning Google, and now finds that most answers from current AIs are pretty reliable. His observation about younger people is even more revealing—they use AI in always-on mode, and more and more of their intangible life goes through the AI and no further. “The AIs are becoming the arbiters of truth,” he concludes.

This is one of those plausible scenarios worth keeping on your radar. Scary? Yes. Impossible? Unfortunately, no.

“We are growing extremely powerful systems that we do not fully understand.”⁠

Jack Clark argues for a blend of technological optimism and appropriate fear: scaling keeps unlocking unexpected capabilities (including signs of situational awareness), yet alignment remains hard and success is not guaranteed. He calls for a transparency regime built before the inevitable crisis, grounded in listening to public anxieties (jobs, mental health, misalignment) and publishing real data. The post threads concrete risks—from model sycophancy to AI‑enabled biosecurity bypasses—while insisting that acknowledging the “creature” we’ve built is the only way to tame it.

Julian Schrittwieser examines why public debate keeps missing the exponential curve in AI progress. He highlights METR’s time‑horizon evals and OpenAI’s GDPval to show consistent capability gains across software and broader occupations—and argues that simple extrapolation points to near‑term, real economic impact. A clear, grounded read on where we’re headed.

Over a 12-month period, while delivering higher-quality responses, the median energy consumption and carbon footprint per Gemini Apps text prompt decreased by factors of 33x and 44x, respectively. Based on our recent analysis, we found that our work on efficiency is proving effective and the energy consumed per median prompt is equivalent to watching television for less than nine seconds.

Google seems to be taking a transparent approach to measuring and improving the environmental footprint of its AI models, sharing detailed methodology and impressive efficiency gains. Their commitment to data center optimization and clean energy scaling is a strong signal for the future of sustainable tech. It’s encouraging to see major players publish real numbers and push for accountability as AI adoption accelerates.

Sam Altman argues that the rise of digital superintelligence will be gradual and transformative, with major benefits for science and productivity. However, critics say his vision is overly optimistic, underestimates current technical and social challenges, and glosses over risks like inequality and AI alignment. Others, such as James Pethokoukis, see the essay as a strategic attempt to calm public anxiety, while some urge a more grounded debate about AI’s real-world limitations.

This is one of the most wonderful personal projects I’ve seen in a long time. Designer Jenny Volvoski set out to recreate the covers of books she had read, with a few constraints: green/black/white for the colour, Futura/typewriter/handwriter/(and Caslon Italic) for the text, and scans/drawings/photography for the image. As you can see, the results are stunning.

Anticipatory obedience is a political tragedy. Perhaps rulers did not initially know that citizens were willing to compromise this value or that principle. Perhaps a new regime did not at first have the direct means of influencing citizens one way or another.

This is a very important message on which we should reflect deeply.