Week 21, 2025 - Money for Nothing

(Feel free to scroll through the first part if you get annoyed by Bitcoin - I find it a fascinating invention, a celebration of mathematics, so I’m a bit biased.)

Bitcoin broke its all-time high this week, climbing briefly above $111K, meaning its hypothetical market capitalization went just above 2.2 trillion US dollars. This is counting with all the bitcoin mined until today, including huge amounts that we’re pretty sure are lost forever, or at least until we can feasibly brute-force old wallets. Sure, the weak dollar has a lot to do with this record, but still, I have two interesting reference points to put this 2.2 trillion in context:

• This market cap is around seven times what OpenAI was valued at their latest round, putting Bitcoin above companies like Amazon, Alphabet, Meta, or Tesla based on market capitalization, making it the fourth, just behind Microsoft, NVIDIA, and Apple.
• It’s still "only" around 10% - 15% of gold’s estimated market cap (if you multiply gold’s current price by the sum of the entire world’s above-ground reserves).

On one hand, it’s ludicrous to even compare Bitcoin to well-established, active, profit-making companies, or a centuries-old OG store of value. On the other hand, it’s an asset that was born less than two decades ago, and its use is magnitudes easier and cheaper than gold’s… go figure. Anyway, the amazing thing for me is not that it broke its all-time high again, but that it did it without basically anyone noticing.

Talking about insane amounts of money, OpenAI bought IO, the hardware startup of John Ive, for 6.5 billion dollars. Watching the short film produced by the studio of the Coppolas, every frame radiating the amount of money spent on this, what's essentially an announcement of an upcoming announcement, I can’t help but think that this can’t end well. Radically new, disrupting ideas are coming from new underdogs, not old behemoths. I hope to be proven wrong.

This week, reality influenced my Leadtime.tech article: the situation came up in two discussions where a team was struggling with old software development practices that have multiple opportunities for improvement, and the challenge became from "what to do", to "where and how to start". I wrote about it here.

🤔 Articles that made me think

Your 30/60/90 is too slow for startups

Joshua is coming from a Product background, but his advice is interesting for Engineering too. Essentially, he’s saying that the traditional 3-month onboarding framework is too slow for startups. I’m sure the proposed 3/6/9 is more about look than reality, but his advice is solid: start doing things sooner than you feel comfortable to, and delay your judgments until you have more context. This is a better way to earn trust and contribute successfully in the long term.

Sean Goedecke’s Blog

I rarely recommend an entire website and not just an article, but I’m super excited every time I discover something new that’s full of great content, as is the case with Sean’s blog. I popped in via his recent article about The Valley of Engineering Despair, that all too familiar feeling when we’re after the initial rush of kicking off a project, still some decent amount of work to do before the deadline, and just realizing things are way less straightforward than we thought. His advice is great: 1) expect this phase, as it will definitely happen; 2) calm down when it strikes; 3) address one issue at a time, starting with the scariest; 4) overcommunicate for transparency and potential help.

Digging deeper into his site, there are great articles from business to virtue. He just collected the best ones in a book, which you can even download for free if you can’t afford the on-demand print version. I just impulse-bought one, and if I'm not disappointed, I'll do a review here.

Heavyweight – Season 9 Episode 3

My favourite non-engineering and non-leadership podcast, the emotionally moving and absurdly funny Heavyweight, has finally found a new home at Pushkin Industries, after being slowly suffocated by Spotify, following the doomed acquisition of Gimlet Media back in 2019. They are just warming up with shorter stories, but it’s already awesome to hear new material from those familiar voices.

🎬 Something Cool: Google Veo 3

I haven’t used the latest version of Veo, but these examples are mind-blowing. László Gaál, who nailed that fake Porsche advertisement, is playing with the new model:

What fascinates me is that these videos are the results of simple, short prompts like “a man doing stand up comedy in a small venue tells a joke (include the joke in the dialogue)” — for some reason, I always assumed it takes paragraphs of carefully crafted text to make sure every detail is right, but no, we’re already past that phase.

Somebody joked on Reddit, saying, “My grandma isn't ready for this” - heck, I’m not sure I am! And as they say, these technologies are only going to improve from here.

That’s it for today, get out of the Valley of Despair this weekend,

Péter