Writ very small, though. You can easily fit a dozen bikes into the space of one parking spot, if not more (double-decker racks exist!), and it is a lot easier to contrive a spot for your bike in the absence of bike racks than it is to park a car when there's no parking.
Heck—if you have a car & your building doesn't have parking, you're basically screwed. If you have a bike & it doesn't have a bike rack, you can just carry it up & put it on your balcony. At that point, I don't think you can really compare the two.
That's a common narrative in popular culture (especially since the publication of Jonathan Haidt's The Anxious Generation), but it doesn't really bear out in data. Smartphones don't really have a discernible impact on mental health at a population level.
The idea is that teen mental health got dramatically worse in the early 2010s at the same time as social media began to become ubiquitous, but this is likely a coincidence. The underlying metrics we're tracking here are self-harm hospitalizations, and concerns about teen self-harm were already growing in the early 2000s. This leads to a bunch of new guidance getting published which increases teen mental health screening, tracks mental health status as a cause of injuries, and forces insurance companies to cover associated costs.
It's one of those situations where our stats about a problem increased as we became better at tracking it. Teen suicidality is actually WAY down over the past ~30 years.
Qualitative data is, of course, much harder to work with than hospitalization numbers, but the data we do have suggests a weak correlation, if any, between phone use and poor mental health— see https://pubmed.ncbi.nlm.nih.gov/30944443/, which suggests phones can explain at most 0.4% of variance in well-being among teens. [1]
It feels like common sense that social media is bad for you, and sure, there's plenty of work to be done in understanding how and why social media can cause harm. But the idea that there's some big crisis just doesn't pan out.
[1]: In fairness, Haidt published a response to this article featuring a new, bespoke set of controls for the data. His analysis suggests that the impact of social media use on mental health is nearly twice as large as that of being sexually assaulted and four times larger than hard drug use (which itself has a slightly larger effect size than wearing glasses). Personally, I don't find these conclusions plausible at all. Maybe Haidt's been p-hacking, or maybe the data set is worthless. I couldn't say. https://www.sciencedirect.com/science/article/pii/S000169182...
Everyone wants a scheme for organizing notes, and let me say: Just do it chronologically. It's fine. Taking notes is more important than reviewing them, and the older a note is, the less likely you are to need it.
You can add a layer of structure by periodically reviewing recent entries & updating an index, but I've never felt the need to.
I actually might want to subscribe to your newsletter, provided I read & enjoy your article. So why does the pop-up always interrupt me before the page has even finished loading?
If you inset an unobtrusive newsletter button 60% of the way through the article, perhaps I'll actually click it (or, more realistically, follow your RSS feed).
It feels like the type of infrastructure envy that leads engineers to spin up a k8s cluster to serve a static website. If you're a publishing academic who needs to ingest a lot of material across a variety of subject areas with an eye to producing more, you will need to take a lot of research notes and you will inevitably organize them one way or another. If you are a working software developer, your needs are very different. I think most of the HN-blog-ecosystem-adjacent zettelkastens exist mostly for the fun of playing with information.
"Can't" means it would be bad for business. I think consumers are a lot less turned off by the idea of a OneDrive subscription than a Windows subscription. Better to stitch little services like OneDrive and Copilot into every part of the system and cajole people into paying for those instead.
It bothered me in theory, but when I started writing lua it quickly became clear that it really doesn't matter. It's just another quirk, like significant whitespace: I don't prefer it, but it's so far down my list of language priorities that it basically doesn't matter.
I wouldn't expect so. These machines have been trained on natural language, after all. They see the world through an anthropomorphic lens. IME & from what I've heard, they struggle with inexpressive code in much the same way humans do.
One truism about coding agents is that they struggle to work with bad code. Code quality matters as much as always, the experts say, and AI agents (left unfettered) produce bad code at an unprecedented rate. That's why good practices matter so much! If you use specs and test it like so and blah blah blah, that makes it all sustainable. And if anyone knows how to do it right, presumably it's Anthropic.
This codebase has existed for maybe 18 months, written by THE experts on agentic coding. If it is already unintelligible, that bodes poorly for how much it is possible to "accelerate" coding without taking on substantial technical debt.
i think you are conflating anthropic (the startup) with claude code (the leaked source of one of said startup's products)
i.e., the claude code codebase doesn't need to be good right now [^1] — so i don't think the assumption that this is an exemplary product / artifact of expert agentic coding actually holds up here specifically
[^1]: the startup graveyard is full of dead startups with good code
I'm not taking a "zero technical debt" stance. Every successful startup eventually ends up treating their 1.0 codebase as a legacy burden. That's the cost of doing business, but it is a real cost. If Anthropic could have kept their code quality high with a few clever tricks and a few extra tokens, surely they would have.
So I take this whole episode as confirmation that code which is subject to high-velocity low-supervision work by agents decays quickly even under ideal circumstances.
That's actually a good point, haha. The worst-case scenario of computers being thin clients for other people's servers dissolves when you realize that chromium/electron IS, nominally, a thin client for HTTP servers, and it'll gladly eat up as much memory as you throw at it. In the long term, modulo the current RAM shortage, it turns out it's cheaper to ship beefy hardware than it is to ship lean software.
Heck—if you have a car & your building doesn't have parking, you're basically screwed. If you have a bike & it doesn't have a bike rack, you can just carry it up & put it on your balcony. At that point, I don't think you can really compare the two.
reply