mplanchard

AI seems to be a good tool for people who are already good at learning and critical thinking to use to accelerate their learning. Research on broader populations so far is not so hopeful, with most people using it to avoid learning. This is unsurprising to me, since people are often quite resistant to learning.

As to your point, literacy rates have been declining since the 80s. 40, going on 50 years is plenty of time to study a trend and draw conclusions.

Can you point me to any time in history where low literacy led to better outcomes for the average person than high literacy? Or even a study that looks solely at modern times and doesn’t manage to correlate literacy with positive social achievement?

But even if you can’t accept a historical argument due to the spans of time being incomparable, I personally believe from a more normative perspective that literacy is a foundational and fundamental good for humanity. But, if you can’t accept that argument either, consider how much of our ability to interact with and understand our own society requires literacy: imagine trying to sign up for healthcare, understand your taxes, weigh the risks of a surgical procedure, or figure out who to vote for if you either can’t read, can’t understand nuance, or are constantly missing words due to a lacking vocabulary.

And of course, being less literate means you are more likely to be convinced by people whose rhetoric is convincing, whether or not their ideas are correct and whether or not they are actually trying to fleece you.

Do you speak more than one language? I think that is also a telling experience. It would be extraordinarily difficult for me to live in a country where my second language is the only one. Sure, I can follow along with a conversation well enough, but dealing with government bureaucracy, forms, healthcare, and politics without the ability to be certain in my interpretation of the text would be a nightmare.

The irony of people sitting here on this all-text discussion site, responding to a textual article about the effect of technology on literacy, responding with “well, maybe literacy is actually not so important,” is so utterly ridiculous that I hope it’s satire and I’m just missing the joke.

This is in fact one of the primary points of TFA and the book it recommends: that the degradation of literacy, public discourse, and general intelligence that started with the television (and was well-described in Amusing Ourselves to Death, referenced in TFA, and which the title pays homage to) has only continued and, if anything, accelerated with the introduction of the smartphone and modern social media.

It is less annoying, at least to me. I can read a semi-promotional post. I just bounce right off LLM writing

This is such a classic HN take, thinking you know better without any real knowledge of the situation.

Here’s a real-life blind person who strongly disagrees with you: https://sightlessscribbles.com/posts/disability-dongle/

iphone, linux (ubuntu, nixos), mac, firefox.

I have a shared vault with my partner, which was easy to transfer and re-share. She’s tech savvy, so no issues there, but it didn’t seem any more complicated to me than setting up 1password was.

For other family members, I have tried for years to get them onto macs, mostly successfully, and so for them I just recommend apple’s builtin thing anyway.

I switched to Proton Pass yesterday, since I’m already a subscriber to Proton Mail. Setup and import from 1PW was easy, took maybe 15 minutes tops. So far it has been working satisfactorily enough.

Apart from everything about DHH, I personally feel like the choice to invest so much money in Omarchy, of all things, shows a lack of good technical sense from the folks at 1pw. If you’re doing charitable giving, why not give to an established product that will be around in 10 years time, or one that provides core technology the other linux distros depend on? I don’t want the money I spend on 1pw going to some poorly chosen bet on a vibecoded distro, just because it’s in the SV hype limelight at the moment.

Oh whoops I ceded my livelihood to an opaque corporate subscription service and they banned me

I literally only have Firefox installed and cannot remember the last time I encountered a broke web page.

Too much RAM use is rare for me, but when it happens I just open the tab unloading page and unload all the tabs. Usually it’s Jira’s fault.

I’m sorry, but this is not my experience at all. The process of installing and using Firefox for a non-technical user is no different from Chrome: both give you immediate access to search, both offer to sign you in for cross-computer sync, and… that’s pretty much it.

I have gotten most of my family and friends to switch to Firefox, and I have never gotten a Firefox-specific complaint or question that I can remember.

Seriously, I’m with the person who says people hating on Firefox are often just trying to justify sticking with Chrome due to the inertia.

I’ve been exclusively using Firefox for many years. I don’t even have a Chromium browser installed! (I just nix-shell it if I need to test anything.)

I love the raft of new features that have been coming through recently! I heavily use tab groups, profiles, split view, and the baked in VPN.

The loud ones are always the complainers, which is a problem for adoption, which is exactly what TFA is getting at.

The nice thing about wayland is you can easily combine a compositor/window manager with a wayland desktop shell and get a lot of the stuff you used to have to fiddle with when using like i3 for free. For example, I use niri with the “dank material shell” desktop shell and get status bar, notifications, clock, suspend/resume, etc. all “out of the box.”

This may be unpopular? But Signal > iMessage > WhatsApp, is my personal ranking. Usually it’s possible to get almost anyone on one of those, and while WhatsApp is far from perfect, it’s better than SMS or another totally unencrypted service.

As a certified Meta hater, I would love a better third option, if anyone knows of one, but WhatsApp has such strong network effects, and it seems like a decent way to moderately* improve the security posture of non-tech-savvy family.

*The thing where they store the encryption key with the backup is terrible, and can make the improvement essentially negligible in cases of govt subpoena or similar, but even so, using WhatsApp would still prevent stuff like this SnapChat story

I haven’t! It seems akin to ACP, is that accurate? Do you recommend it?

Mostly same as with any treesitter mode: it’s easier to build structurally-aware features, and it automatically supports any minor mode that’s enabled by treesitter (examples from my own config include structural movement, folding, selection, indentation guides, etc.). So, rather than every language having to reinvent the wheel, and every user having to install six different language-specific packages to get feature parity, more stuff just works.

You mention eg poly mode, which has always been extremely finicky for me. Syntax highlighting in treesitter modes is both more capable and more consistent.

One of my coworkers uses “load-bearing” ironically constantly now, it’s very funny

One option would be to read the section of the linked article talking about features: https://rahuljuliato.com/posts/markdown-ts-mode-emacs-31#a-q...

Thanks for posting the ghostel link, I hadn’t heard of it. Excited to try it out and see how it compares to vterm

Seconded, agent-shell is my favorite LLM interface bar none

I think eglot often just works now. I recently started doing c++ stuff, and it just detected clangd and I was off to the races with no config. Same with rust-analyzer for rust projects.

Other folks talking about how rolls are measured, but I have family members who I can’t help but notice when they come to visit use an absurd amount of toilet paper. I think that, rather than using a square or two, they just pull off a bunch and wad it up.

It took a pretty long time for me (like 15 seconds)

I don’t vibe code, or use AI much when programming, but if I want to experiment with it, I’d like to be able to track my code in the code hosting service I’m paying for, without my repos being subjected to arbitrary scrutiny or ideological purity tests.

It would be different if we were talking about a service I did not pay for: in that case, they can do whatever they like, and I have little to no right to complain.

Big, medium, and startups. Big and medium definitely biased more towards leetcode, but I also got it at around a quarter of startups I applied to.

I would. Just finished up a job search recently for staff+, and still got leetcode interviews for most companies.

I think jyx was talking about the throw/try/catch mechanism[0] used for error handling, in contrast to error handling in e.g. Rust, where the idiomatic way of handling an error is to return a Result enum, containing either a success value or an error type. The latter allows callers to know that a function can error, and what kind of error it can return. Exception-based error handling, on the other hand, means you cannot tell by way of the type system whether or what errors a function can throw.

[0]: https://github.com/abhiramasonny/jaithon/blob/main/LANGUAGE....

Ha! A friend and I had this idea ages ago when living in Austin. I actually have the domain “walkdarkly.com” registered for it, in case I ever got around to it. If somebody actually doing something with it wants the domain, I’m happy to talk about transferring it

Teen-ager and some other unusual hyphenations are New Yorker style, yes

It’s worth traveling to see totality if you can swing it