jmbwell

I’m so almost here. The thing holding me back is projects that don’t do their own migrations reliably. Through no fault of their own, perhaps, though at this point I would argue LLMs should eliminate any good reason not to have alembic integrated or something. And even Home Assistant is bizarrely averse to fully automated system wide updates. Updating system and core and addons all independently is bonkers. But yes, the simplest implementation is often the best

I just caught myself thinking Pantos shouldn’t have answered the phone. Not in a blame the victim way, more in that never answering the phone is just good opsec at this point. The phone, the door, just don’t talk to a cop without a lawyer. They don’t come to you to be helpful to -you-.

Except in this case in a broader sense, we know about something we wouldn’t have otherwise

Who benefits from all this brilliant deal making who doesn’t already have plenty of money? If we are going to invent money out of paperwork maneuvers, you’d think we could invent a way to fund healthcare and schools.

I might argue that generating and decoding an actual NTSC signal, as the OP project does, would be true in ways that a generative model based on all of that would not be.

I sincerely appreciate this fidelity to fidelity

One of my favorite artifacts of the pre-platform era of the web. Valuable stuff presented clearly … all clearly handmade with care. Not a monetized journey conversion experience in sight

Exploring and at times debating is how we figure things out and share ideas.

What’s pointless is doing so in pursuit of winning rather than understanding

There’s the other direction too. You only get a couple toggles, and something you actually need is behind both, so you can’t not get all notifications

I find this really effective. Also, “ask me questions about this one at a time until I say to stop”

I’ve also moved all mine to Europe. There are ample alternatives to us-based commercial cloud.

The regulatory environment is different, so it’s worth understanding the ramifications as far as what’s expected of you if you’re operating in a different jurisdiction. It’s nothing that can’t be handled, but some may find they have to care about things they haven’t before

It’s a great exercise for shoring up independence from extractive providers

Maybe I should have AI write up an article too. Honestly, it’s not just rare, it quietly matters

Indeed. Understand the principles, you can work with just about any tool

I’ve been finding it difficult to hand things over to an LLM completely, and I wonder if this is part of it. I’ll let it help me organize at the beginning, and then I’ll have it come through and refactor or review, but the crafting part is where I want to spend my time. I love hitting tab but it seems like every time I do, I get this sensation like I’ve sort of time-warped into the future by a few seconds, and I wonder what I’d have written if the LLM hadn’t done it for me. At which point I’ll never know.

Great for generating volumes of output. Less great for going into a flow state and coming out with something that looks like I made it, something that I see my hand in

And yeah, maybe it’s because I never quite get into a flow state when I do it this way. Hmm

The language of drama and import without meaningful substance. Words statistically likely to be used in a segue, regardless of the preceding or subsequent point. Particularly effective when it seems like you’re getting let in on a secret. Really fatiguing to read

A writing teacher once excoriated me for saying that something was important. “Don’t tell me it’s important, show me, and let me decide, and if you do your job I’ll agree”

I don’t know how a completion can tell when it needs to do this. Mostly so far it doesn’t seem capable

It does seem like an LLM’s ability to see a constraint and just say “I’ll write a quick helper to work around it” kinda wrecks some older-world assumptions. We know how to deal with remote human attackers, remote bot attackers, and to some extent local human attackers, but local self-coding bot attackers lately needs more attention than it used to. It’s not even the same category as malware

I’ve been guilty myself of building containers where everything runs as root on the assumption that the container was the relevant domain

If LLMs are involved, I can’t tell whether OS level security is suddenly more relevant, or suddenly utterly obsolete

I understood it to mean, GitHub is being crushed by LLM/AI/Agentic code review and submission, not GitHub’s code itself

What I’m not seeing here but I am seeing with the Linux kernel is, most of the automatically submitted code is irrelevant or not useful

(Maybe that’s what you were getting at, apologies)

These days, I’ll take Mad magazine

I struggle not to have a cynical take these days. Of course he cared about the ROI. The ROI is access to an underserved market, a halo effect, a new community of adherents, a new reason for customers to cross the moat into the ecosystem… a modest investment with a durable long term return in multiple categories.

I appreciate that it’s a win-win for Apple and for its customers, and I firmly believe that accessibility features serve everyone eventually. I’m glad that there are some billionaires who also see it that way.

I guess I just wish we didn’t have to rely on rare cases of billionaires finding it in their own best interest to happen to serve the rest of us. Especially when the actual accessibility work and everything else is actually done by a whole class of people that never make headlines just for leaving their jobs and being replaced.

“Brand” and “branding” is arguably the most important thing -not- mentioned in the article. The commercial incentives to differentiate are powerful enough to kick a lot of UX out of the way.

Now that all we do is “experience” a “journey,” it’s more about the user doing what the app wants instead of the other way around

macOS is slightly more consistent among apps that use system controls, but the more custom the app, or the more React Native or Electron it is, the less predictable it is

Infuriatingly, some apps try to be smart — only one line, return submits; more than one line, return is a new line, and command-return submits; but command-return on just one line beeps an error.

Years of muscle memory are useless, so now I’m reaching for the mouse when I need to be clear about my intent

So much is solved when developers just use the provided UI controls, so much well-studied and carefully implemented behavior comes for free

I get it. I also know customer service is a pretty low paying job for something that involves being yelled at all day

I get as frustrated as anyone, but it’s not the fault of the person whose job it is to take my call.

I remember once on the phone with Comcast I just explained the situation and jokingly said look, if it helps, feel free to tell ‘em I’m yelling and screaming. The guy laughed. An engineer called me an hour later with a firmware update for my modem.

Sometimes there’s no winning. But sometimes it helps if you can put people on your team

Yes, and this has been the case for years. Cnet, ZDnet, PCmag have been user-hostile since long before AI summaries. Pop-ups, “before you go,” back jacking, all the worst.

The Verge is a surprise because it is relatively new and was relatively free of this crap for a long time.

They’re all just empty brands now. They totally caved to advertisers, and now only advertisers care about them.

I dare say AI’s popularity is a symptom of all this more than a cause.

This. The truth is still putting on its shoes and all that

I note that Ken Fisher did post an editor’s note, Benj did publicly own up to it, and all of this was mentioned in the article.

In writing code, as in writing poetry, the mechanical labor is 5% writing, 45% editing, and 50% reading. But the only thing that makes it yours is you.

Color schemes voluntarily added by the user to an app like vim, great.

All the more reason for developers to keep the app itself responsive to the user’s environment by default.

Don’t bake in elaborate visual choices. It’s a usability thing first and a style thing somewhere much farther down the list.

Keep it simple from the factory. Don’t get in the way of customization. Let the user’s environment do the work of adapting it to the user.

I think fantastical isn’t totally inaccurate, and I’m not being snarky (for once). The personal observations and sometimes colorful language has been something I like about Ars. Benj in particular, with his warm tributes to BBSes. Or Jim Salter’s very human networking articles. The best stuff on Ars is both technically sound and rich with human experience. “Fantastical” taken to mean something like, capturing the thrills and aspirations that emerge from our contact with technology, seems fair I think.

I’ll be interested in finding out more about just what the hell happened here. I hardly think of Benj or Kyle as AI cowboy hacks, something doesn’t add up

And then a handful of companies can offer a service to let advertisers punch through the plugins. And then another plugin could block that!

Thing is you’re probably right. The modern web is made of middlemen inserting themselves into user experiences to divert and extract revenue from the primary stream between consumer and producer. There’s always room for another layer.

I think targeting web sites was the right move because it was the web sites who were doing all the tracking.

Of course now we also have browsers to worry about as well, being products of the same ad companies that were clogging up the web sites in the first place.

But if cookie laws pushed data collecting web sites to malicious compliance, surely similar laws would do the same to (also data collecting!) browser providers. I’d prefer to avoid inviting browsers to add another layer of bullshit. And there’s no reason it would make web sites behave differently… if I’m a web site bound to comply with laws, I’m probably going to cover my own ass and keep doing what I’m doing without assuming the browser will handle it. Rendering the browser controls redundant and ineffective.

If we want to look for core flaws, look at allowing a handful of giant companies to control the market for personal data — or to traffic in personal data at all.

Ad companies have convinced the whole economic system of the Internet that they are inevitable and essential. They are neither. But we won’t fix that either.

The solution is to get off the damn internet, but short of doing that, I’ll prefer to keep my options open to disable telemetry on my own terms.

Here’s something I would like, though: total sandboxing per web site. Let every domain be alone in its own room of cookies and telemetry. Let it think I only ever visit that site, and optionally always for the first time. I shouldn’t have to blow away all my cookies all the time just to keep Facebook from following me all over the web.

I tried the same questions with my own account. I was surprised at how much it was able to synthesize that wasn't completely off-base.

With these sample questions, there wasn't much to learn, and it gave me relatively thoughtful-seeming responses. Nothing alarming -- I would expect it to recall things I've discussed with it, and it's very good at organizing things, so it's not a surprise that it did a good job at organizing a profile of me based on my interactions.

I would be curious how crafting the questions could yield unexpected or misleading results, though. I can imagine asking the same questions in different ways that might be designed to generate an answer in support of taking particular action. If I wanted to arrest me at the border, for example, I could probably ask questions in such a way that the answers would make me look arrest-able easily.

So this is my concern with ChatGPT -- not that it will reveal some unseen truth about me, but rather that it is trivial to manipulate it into "revealing" something false, especially as people consider it to be more capable and faithful than an elaborate sorting algorithm could ever be.

- Meds are just one of the tools available, just one part of a holistic approach that includes other accommodations, practices, and support from the people around you

- Not everyone experiences these things in the same way

- Your goals are for you to set; if incorporating meds into your plans doesn’t help you reach your goals, fine. But if meds help you unlock goals you might not be able to access otherwise, maybe they’re worth considering.

- The vast majority of professionals really do want to help you reach your goals; most psychiatrists (for the meds) and psychologists (for your cognitive health) are going to be more valuable in terms of perspective than an Internet thread :D