Posts

I’ve made my share of bad predictions. I’ve had plenty of good ones too. But being right often doesn’t matter, because few people will remember if you were right or wrong, and when it comes to placing bets on your predictions, timing is incredibly difficult to nail down. Some of my best predictions include my general bullishness on software and the internet since the 90s (when the dot-com bubble saw the NASDAQ rise over 400% from 1995-20001), betting on Airbnb (now valued at $86B+), and Bitcoin (up over 30,000% in the last decade). In the case of Bitcoin, I was right about the bet, but got the timing wrong. Bitcoin went from $0.09 in 2010 to nearly $69,000 in November 2021. A $100 buy in 2010 would have been worth over $76 million at the top.

Once upon a time, the US government enforced antitrust legislation. One of the most famous examples of this is the case of Standard Oil, which remains the textbook example of monopoly breakup taught in economics and law schools worldwide. To understand what happened to Standard Oil, we need to take a short walk back in time through history. The year is 1890, and the US Congress is a well-functioning body with rational officials whose brains still function at a high enough level to pass sensible legislation.

I had the recent pleasure of speaking with someone from a big tech company about their interview process. They wanted to put me through the funnel to run the gauntlet and see if I “had what it takes” to make it to the offer stage. I declined, but (against better judgement) I gave them 30 minutes to pitch their company to me anyway.

Photo by Denys Nevozhai Cars kill 1.3 million people every year, eat up huge chunks of our cities, and pump out a depressing amount of pollution. We’ve built everyday life around them anyway.

Humans are pretty bad at intuiting about exponential growth. Whether it’s pollution, population growth, or the way a virus spreads, our brains struggle to grasp the impact of such growth.

You’ve probably heard about the “labour shortage” everywhere. But what’s really happening in the job market? People point to early retirements, COVID deaths, or stimulus checks, but the numbers don’t neatly support any one of those stories. Let’s explore what the numbers actually show about the current state of the labour market.

We’re neck-deep in self-promotion and personal branding. You can see it most clearly in the rise of “influencers”—a term that says plenty on its own. A lot of them aren’t influencing much of anything; they’re just calling attention to themselves. Consider this: when someone uses carefully curated photos of themselves on vacation to promote products, are they truly influencing, or are they themselves being influenced by sponsors? In many cases, they’re posting ads with their own face attached. The platforms and brands they represent are often the true beneficiaries of this arrangement.

I’ve noticed a trend amongst startups (particularly those in the financial-ish space) which amounts to some variation of “hey guys, maybe we can get customers if we use memes?”. This is not dissimilar to the “viral” craze of the early 2010s in startup land, where a number of companies raised mountains of VC cash entirely on some silly app gaining millions of users overnight. One of my favourite examples is the Yo app.

I’ve been thinking about the relationship between friendship and therapy in our modern world. It’s become increasingly common to suggest therapy as the default solution for nearly every emotional challenge, as if talking to a professional is the obvious answer whenever life starts to hurt. Therapy can be valuable and sometimes necessary, especially for people dealing with clinical conditions or trauma. But I think we’ve gotten too used to professionalizing emotional support and forgetting what good friends are for.

Software engineers are likely familiar with the CAP theorem, coined by Eric Brewer. In essence, the CAP theorem states that distributed databases can have at most 2 of 3 attributes: consistency, availability, or partition tolerance (fault tolerance). In other words, you can have a database that is consistent and highly available, but not fault tolerant. Or, you can have one that’s fault tolerant and highly available, but isn’t always consistent (especially during network failures).