Siddhant Pusdekar talking about how lifelike biochemistry continued to unfold in sterilized soil for six years, pointing to a metabolic theory for how biology began.

For 15 years, Sébastien Fontaine has been trying to kill dirt. The biochemist, who runs a lab at the French National Institute for Agriculture, Food, and Environment, wanted to know how much carbon is released by soil — just dirt alone, completely devoid of life. His team sealed dirt into jars and blasted them with sterilizing gamma radiation. Then they waited for the carbon dioxide released by the soil — a sign of ongoing microbial respiration — to drop.

They waited, and waited, and waited some more: weeks, then months. Under a microscope, the irradiated soil showed no signs of life, but it continued to emit carbon dioxide. The soil wouldn’t stop breathing.

Fontaine’s lab repeated the experiments and produced the same results. Finally, convinced that they weren’t dealing with an artifact of the experimental setup, they set out to find the source of breath in dead soil.

Now, Fontaine and his colleagues have reported that their soil samples continued to consume oxygen and spew carbon dioxide for six years. In a 2025 paper in Science Advances, they proposed that a metabolic process that powers much of life is also possible outside living cells. Their experiments point to how it could work in dirt, absent the living proteins that would typically organize it. If they’re right, some biochemical reactions, such as those that release the energy of carbon-rich sugar molecules, may not be unique to living things. Such reactions — known as metabolism when performed by cells — could even predate life on Earth, Fontaine said.

The experiments show “what happens to biomolecules when they’re left to their own devices,” said Joseph Moran, an organic chemist at the University of Ottawa who was not involved with the research. They’re finding that the chemistry of life is not exclusive to life, he added. “It’s the chemistry of geology.”

Maciej Cieglowski’s wonderful, and at times hilarious, talk on super intelligence which he gave in 2016. Yes 2016. It’s a long read, but worth every second of your time.

On intelligence level:

Our intelligence level, cognitive speed, set of biases and so on is not predetermined, but an artifact of our evolutionary history. 

In particular, there’s no physical law that puts a cap on intelligence at the level of human beings.

A good way to think of this is by looking what happens when the natural world tries to maximize for speed.

If you encountered a cheetah in pre-industrial times (and survived the meeting), you might think it was impossible for anything to go faster.

But of course we know that there are all kinds of configurations of matter, like a motorcycle, that are faster than a cheetah and even look a little bit cooler. 

But there’s no direct evolutionary pathway to the motorcycle. Evolution had to first make human beings, who then build all kinds of useful stuff.

So analogously, there may be minds that are vastly smarter than our own, but which are just not accessible to evolution on Earth. It’s possible that we could build them, or invent the machines that can invent the machines that can build them.

There’s likely to be some natural limit on intelligence, but there’s no a priori reason to think that we’re anywhere near it. Maybe the smartest a mind can be is twice as smart as people, maybe it’s sixty thousand times as smart. 

That’s an empirical question that we don’t know how to answer.

On general intelligence:

The concept of “general intelligence” in AI is famously slippery. Depending on the context, it can mean human-like reasoning ability, or skill at AI design, or the ability to understand and model human behavior, or proficiency with language, or the capacity to make correct predictions about the future.

What I find particularly suspect is the idea that “intelligence” is like CPU speed, in that any sufficiently smart entity can emulate less intelligent beings (like its human creators) no matter how different their mental architecture.

With no way to define intelligence (except just pointing to ourselves), we don’t even know if it’s a quantity that can be maximized. For all we know, human-level intelligence could be a tradeoff. Maybe any entity significantly smarter than a human being would be crippled by existential despair, or spend all its time in Buddha-like contemplation. 

Or maybe it would become obsessed with the risk of hyperintelligence, and spend all its time blogging about that.

On AI cosplay:

The most harmful social effect of AI anxiety is something I call AI cosplay. People who are genuinely persuaded that AI is real and imminent begin behaving like their fantasy of what a hyperintelligent AI would do.

In his book, Bostrom lists six things an AI would have to master to take over the world:

• Intelligence Amplification
• Strategizing
• Social manipulation
• Hacking 
• Technology research 
• Economic productivity

If you look at AI believers in Silicon Valley, this is the quasi-sociopathic checklist they themselves seem to be working from. 

Sam Altman, the man who runs YCombinator, is my favorite example of this archetype. He seems entranced by the idea of reinventing the world from scratch, maximizing impact and personal productivity. He has assigned teams to work on reinventing cities, and is doing secret behind-the-scenes political work to swing the election. 

Such skull-and-dagger behavior by the tech elite is going to provoke a backlash by non-technical people who don’t like to be manipulated. You can’t tug on the levers of power indefinitely before it starts to annoy other people in your democratic society.

I’ve even seen people in the so-called rationalist community refer to people who they don’t think are effective as ‘Non Player Characters’, or NPCs, a term borrowed from video games. This is a horrible way to look at the world.

So I work in an industry where the self-professed rationalists are the craziest ones of all. It’s getting me down.

Aaron Levie talking to Casey Newton about what kind of jobs he expects in the future dominated by AI coding assistants.

Newton: It’s the sort of thing I love to hear. I would love to not live through a massive disruption where we see super high unemployment. I also can’t help but note we saw almost 46,000 tech layoffs announced in March alone, with AI sometimes cited as a potential cause. So would you put a number to it? On the software engineering front, do you think in three years we have about as many software engineers as we have today, or more? Or do you think there’s a bigger shift?

Levie: I think we’re going to have more. I don’t want to be unsympathetic to people who really will face these changes. But big picture: if you were a CS grad of the past two decades from a top 25-to-50 CS school, by and large you were trying to go to a tech company — in Silicon Valley or a couple other places. So most of the software talent in the world of this cohort ended up building software for consumers. We were building ad apps, ride-sharing, enterprise software (thank God). We’ve accumulated a lot of engineers on that kind of work, and some of those companies have over-hired.

Who’s the loser of that equation? Every other company on the planet, because they couldn’t compete with Google and Facebook and Microsoft for that top engineer. They couldn’t automate things in the life sciences process, or the supply chain, or automotive AI systems. I don’t know how much software you’ve used from companies that aren’t in the Valley, but if you log into your bank and you’re happy, you’re a totally rare person. If you look at most car console designs of any car that’s not from two companies, you can imagine how unusable these systems are. That correlates to the fact that those companies couldn’t overstaff with all the top engineers and designers.

Now what happens? All of a sudden, what was maybe a 30- or 50-engineer problem previously, Claude Code and Codex come in, and now it’s a 5- or 10-engineer problem. For the first time ever, those companies are able to take on work that wasn’t possible before. They can bring automation to all the systems and workflows they couldn’t have afforded or justified.

So in some cases of tech, you’ll see a temporary dislocation. At the exact same time, the thing you should be tracking is the number of engineering jobs opening up at traditionally non-Silicon Valley tech companies — small businesses, consulting firms, life sciences, manufacturing.

And as a force multiplier, you’re going to have a number of new types of engineering jobs where the job is entirely about how to deploy agents inside the firm to automate work. I did this for fun just to make sure I wasn’t full of shit. If you go to the Eli Lilly careers page, as one does, they have this job title called “lab automation software adviser.” That person is an engineer whose job it is to bring automation through AI to the lab process.

Think about how many hundreds of thousands or millions of jobs will look like that in the future. My job is to take the innovation coming from AI-land and apply it to this particular business process in my organization. You’re kind of like an FDE — a forward-deployed engineer — but for that company. Those will be the people who would have gone to Meta or Google five years ago. They’re going to now work in pharma, banking, manufacturing. And those are actually incredibly stimulating jobs. You’re not just building an app, you’re automating drug discovery.

For the sake of my career continuity, I really hope that these forward deployed engineers come from Indian IT industry.

This informative post by Nick Mark explaining what VO₂ max is and how it varies from humans, sled dogs, pronghorn, bats, hummingbirds, and bumblebees.

VO₂max – the maximum rate of oxygen consumption during peak exercise – is one of the most informative numbers in physiology. Among athletes, physiologists, and intensivists alike, it functions as a kind of summary statistic for the entire oxygen delivery cascade: how well the lungs extract oxygen, how efficiently hemoglobin carries it, how powerfully the heart pumps it, how densely the capillaries deliver it, and how effectively mitochondria consume it. Each step is potentially rate-limiting, and VO₂max tells you how well the entire oxygen cascade functions. VO₂max also happens to be one of the strongest independent predictors of all-cause mortality.

VO₂max is best understood not as a single trait, but as the product of a chain, as described by the Fick equation:

VO₂max = Cardiac Output × Arteriovenous O₂ Difference

Or, expanding the Fick equation more fully: pulmonary ventilation × extraction efficiency × hemoglobin concentration × cardiac output (stroke volume × heart rate) × capillary density × mitochondrial oxidative capacity. Each of these steps can be a bottleneck. In the short term, training, and in the long term evolution, can widen any these bottlenecks.

This comment by gopalv from Hacker News:

If you manage 500+ people organization, most of the headaches with agents already exists with you – you set directions, ask people to go run fast in those directions, check in frequently and course correct on results without actually understanding those people do.

Those aren’t the deal breakers.

They entirely rely on the competence of the folks they hired and cross-match enforcers with the drivers they have – they deal with fallible people on both sides of that.

The fundamental difference is that the humans are good consequence predictors, have built up reputations they are not willing to trash, can say no to things and in general don’t want to go jail.

AI tools look like that, but don’t have any of the useful conflict which came for free with employing humans.

It also doesn’t have any useless conflict, but not all conflict between what I say and what someone is willing to do is bad conflict.

Brian Potter explaining how Helium is produced and where it is used.

Helium is the second lightest element in the periodic table (after hydrogen), and the second most common element in the universe (also after hydrogen). But while helium is very common on a cosmic scale, here on earth it’s not so easy to get. Because helium is so light, it rises to the very top of the atmosphere, where it eventually escapes into space. So essentially all helium used by modern civilization comes from underground.

Helium is produced via the radioactive decay of elements like uranium and thorium, and it collects in underground pockets of natural gas. This source of helium was first discovered in the US in 1903, when a natural gas well in Kansas produced a geyser of gas that refused to burn. Scientists at the University of Kansas eventually determined that this was due to the presence of helium. Like petroleum, helium has collected in these pockets over the course of millions of years, and thus (like petroleum) there’s a limited supply of underground helium that can be extracted. As with petroleum, people are often worried that we’re running out of it.

Because helium is a byproduct of natural gas extraction, and because only some natural gas fields have helium in appreciable quantities, a small number of countries are responsible for the world’s supply of helium. The US and Qatar together produce around 2/3rds of the world’s helium supply. Russia, Algeria, Canada, China, and Poland produce most of the remaining balance.

Much like the movie Surrogates starring Bruce Willis, Om Malik reflects on the idea of creating a digital version/twin of yourself to interact with your readers (or fans).

The more I think about it, the more I realize this is the ultimate expression of what began in the social media era, when media manipulation became the primary currency instead of authenticity. We all created curated, and often false, lifestyles on Instagram.

Social media gave us tools to edit our lives into a highlight reel. Photos of coffee, food, selfies from places you couldn’t afford last year, some pithy comment. It was all one directional. A movie about me, by me, for me to broadcast and you to watch. This is what led to the rise of influencer culture, where anything and everything was for sale. The self first became a gallery, then a reel. It was all passive, beautiful, controlled and fake.

We shared bumper sticker wisdom on Twitter. LinkedIn became a public square to hawk faux expertise. This popsci compression of complex thinking into shareable nuggets, designed for distribution and optimized for engagement, was the next step in the self becoming a product.

The pseudo-conversation twin is the crescendo. The self’s full immersion into illusion is now interactive. It answers questions. It gives the impression of encounter, of dialogue, of relationship. But it is still the same curated self with a conversational interface bolted on. It is as authentic as a Potemkin village. And with every step we have moved further from the actual person. The twin is not a rehearsal. It is the first act of abstraction of ourselves. Reid AI can do the job from a bunker in New Zealand.

[…]

The twin doesn’t just represent you. It restructures how others relate to you. The copy becomes the relationship. Send out the twin, and you have not freed yourself for deeper thinking. You have replaced the possibility of being surprised by another person with the certainty of your own archive.

While reading this post, I also learned about:

Bumper sticker wisdom

are short, punchy, and witty maxims, slogans, or philosophical snippets that are designed to fit on a car’s bumper sticker, but are oversimplified and miss the nuances of the argument.

Potemkin village

is a construction, literal or figurative, that provides a façade to a situation, to make people believe that the situation is better than it actually is.

Bryan Cantrill talking about how using LLMs for writing breaks a social contract.

[…] LLM-generated prose undermines a social contract of sorts: absent LLMs, it is presumed that of the reader and the writer, it is the writer that has undertaken the greater intellectual exertion. (That is, it is more work to write than to read!) For the reader, this is important: should they struggle with an idea, they can reasonably assume that the writer themselves understands it — and it is the least a reader can do to labor to make sense of it.

If, however, prose is LLM-generated, this social contract becomes ripped up: a reader cannot assume that the writer understands their ideas because they might not so much have read the product of the LLM that they tasked to write it. If one is lucky, these are LLM hallucinations: obviously wrong and quickly discarded. If one is unlucky, however, it will be a kind of LLM-induced cognitive dissonance: a puzzle in which pieces don’t fit because there is in fact no puzzle at all. This can leave a reader frustrated: why should they spend more time reading prose than the writer spent writing it?

Derek Sivers:

Forty years ago, a family moved from India to Canada, and raised their children with “Indian values”. When those children visited India last year, the locals laughed at their outdated beliefs. What their family had said were factswere just a perspective from 1980.

[…]

When someone speaks of a place, you have to ask, “When?” Geography is four-dimensional. You can’t know a place – only a place as it was at a time. Where is bound to when. Unless you are in a place right now, you can only speak of it in past-tense.

From the movie The Gods Must Be Crazy:

And here you find civilized man. Civilized man refused to adapt himself to his environment. Instead he adapted his environment to suit him. So he built cities, roads, vehicles, machinery. And he put up power lines to run his labour-saving devices. But he some how didn’t know when to stop. The more he improved his surroundings to make life easier the more complicated he made it. So now his children are sentenced to 10 to 15 years of school, just to learn how to survive in this complex and hazardous habitat they were born into. And civilized man, who refused to adapt to his surroundings now finds he has to adapt and re-adapt every hour of the day to his self-created environment.