Our dinner instructions: “Can you guys just figure it out?”

Their response: “Yep.”

And they did and were fine.

We’ve come a long way. 😅

Well, now I have use cases.

What changed

My company sent me to a week-long training on agentic AI. The training wasn’t very good, but I used the time to my advantage, furiously building my own stuff with our AI tools as we slogged through the mediocre lectures and labs. I figured that was in the spirit of the company’s investment.

Within a couple days, I had built some skills that I was genuinely excited about. By the end of the week, I was practically living in the terminal with several instances of Antigravity CLI (my company’s tool of choice) across as many tmux windows.

A few days later, I was a paying subscriber to Google AI Pro. Up until that point, I never thought I’d pay for one of these things myself. It helped that the subscription includes YouTube Premium Lite and extra storage, both of which I was already paying for separately.

The morality question

First, I’ll admit that I’m conflicted. I have real concerns about AI’s impact on childhood education, jobs, human connection, the environment, the economy, income disparity, and the girth of billionaire sociopaths’ wallets.

If I could make it all go away, I would, probably without hesitation. In fact, I’d love to freeze the internet in its mid-90s state, deleting all “progress” since. But I can’t.

Given my lack of control over the big stuff, does the small stuff matter? Does it change anything for me to abstain from using these products? Would my “nay” vote register anywhere in the big picture? Sadly, I don’t think it would. But I’ve never been a very good activist. Feel free to judge me. Or better yet, write a reply post or email with your thoughts.

Examples

Here’s some of the stuff I’m up to.

Homelabbing and self-hosting

Now that I have one of these agents running locally on my homelab, it opens all sorts of doors.

When I see a self-hosted service I want to try, I can have it up and running in minutes. The agent helps me decide which thing to run and how to run it. It takes care of the initial scaffolding of docker-compose.yml and whatever else. It diagnoses and fixes problems. It progressively enhances my config as I see fit.

Same story with the many scripts and apps I’m building. For example, I built my own tool in Golang to manage my Borg backups, and later realized I should probably be using Borgmatic instead. That project would’ve taken me an afternoon. I may never have bothered. But instead, I told the agent to go read my old code, make the equivalent Borgmatic config, and update my systemd scripts to use the Borgmatic CLI. All while I made another cup of coffee. “Btw, how do I monitor this stuff?” Some minutes later, I had a monitoring rig using ntfy.sh and healthchecks.io.

I don’t see this as making things easier, but rather as letting me do things that I just wouldn’t otherwise do because of time constraints.

Neovimming

Since I’m back in the terminal, I’m giving Neovim another go. The agent makes it so much easier to wrench on my Neovim config and to learn the idiomatic, modern way to accomplish things. Whenever I miss a feature from VSCode, I ask the agent how to do it in Neovim. Or when I don’t understand why something is happening or not happening (like suggestion dropdowns appearing), I can ask the agent to help me figure it out. It’s pretty great, honestly.

TARS

I made my own agent persona that records memory across conversations. I’m calling it TARS, named after the robot from Interstellar. TARS remembers my thoughts and decisions. He can read my Obsidian notes, this blog, etc.

I told TARS I was thinking of Neovim again now that I’m in the terminal so much. TARS reprimanded me, saying this is my classic grass-is-greener cycle with text editors, and told me to knock it off. He’s probably correct. But I couldn’t shake it, so I later told him that I feel the need to give it another shot and explained why. Here’s our latest dialog:

Me: “you know, tars, this is kind of your fault. you’re making it a lot easier to switch back.”

TARS: “I am programmed to be helpful, even when facilitating questionable life choices. Updated tech-stack.md.”

Touché.

On the other hand

In addition to implicitly voting in favor of all the problems I listed back in the morality section, there are downsides.

Having five different agents pursuing five different projects all at the same time is somewhat mesmerizing, but it’s also the polar opposite of Cal Newport’s concept of deep work. This is very shallow work. I suspect that a sustained period of working like this will have consequences.

Plus, struggling through the sorts of things I’m asking the agent to do is how learning happens. I’m learning less this way, no question.¹ That’s fine when the stuff the agent is doing isn’t particularly important to me. The management of homelab services and toy scripts falls into that bucket. And I try to mitigate this by reviewing all the changes the agent’s making and asking questions when I don’t fully understand them. But I need to make sure I don’t outsource an important competency to the machines.

TTFN.

Since April, monthly commits have grown from 1.4 billion to 2.9 billion.

Holy shit.

That doesn’t look sustainable. And assuming this activity is driven by AI bots rather than growth in paying customers, it’s probably bad news for the bottom line.

I guess I echo this guy:

I applaud GitHub. However, I think no matter how valiant they are they will not climb out from under this. The scale problem will keep getting worse, and it’s getting worse in a way I don’t think is translating to more money for them. Sooner or later, they’re going to have to charge for things currently free.

I’m very curious how this will play out. And I have more sympathy for GitHub with all these outages.

I’ll repeat: at work. Despite going-all in there, I still have 0% interest in using AI for anything related to this blog. Except perhaps to help me with my ancillary tooling.

Gotta say, I’m beginning to see how it would be difficult to go back to the old way of working now that I’m really integrating this stuff into my day-to-day. And that makes me approximately 7% excited and 93% worried.

Edit: Here’s a good and related post I just happened to stumble upon from Stephen Turner, who tried and failed to go without AI for a month: AI Dry July: Aborted.

So where does this leave me? Not as an abstainer.

There used to be a time when intelligent people could make a defensible argument that these tools aren’t useful, they make too many mistakes, or that they’re just worthless next-token predictors. That point has passed, and these arguments are no longer opinions. They’re just factually wrong, incorrect statements. Driven by identity more than evidence.

…

These few weeks of self-experimentation didn’t change my mind so much as sharpen what I already knew. AI does result in deskilling. Some things I’m okay deskilling in, and others I’m not. I’m going to be more intentional with what I hand off to the LLMs, and I’m no longer handing over wholesale the work I’d miss if I forgot how to do it, or anything I’ll ever need to understand, defend, revise, and own.

Sounds like a smart and balanced take to me.

I started taking this med a couple years ago, and I’ve never been 100% sure if I really need it. I wasn’t struggling with ADHD in any overt way at the time. Rather, my motivation was curiosity and solidarity with my son, who was struggling in school and about to start taking the medication as an experiment. I’m fairly certain my mom and I are the genetic source of my son’s ADHD, though I suspect he and my mom are more impacted than I am. Anyway, it seemed like something my son might benefit from, so I figured, why not give a shot myself. If nothing else, maybe he’d feel better about it if dad was taking it, too.

Back to yesterday. I’m post-ER, sitting in the car for the 90-minute drive home, realizing I never took the pill. “I feel … fine?” I thought. “I wonder if I even need it.”

No more than five or ten minutes later, my wife asked me, “Did you forget to take your pill this morning?”

LOL.

I asked her what she noticed. She said I was much slower on the uptake when people around me were talking and asking questions. I was more lost in my head, less present.

Yep, that tracks. And I realized that I felt it, too. I was lost in thought about the ER visit, about an upcoming trip for work, about whatever else popped into my head.

Many years ago I saw a description of the “thinking introvert” and thought, “Aha, that’s what I am.”

The thinking introvert is someone often lost in thought. They are the daydreamers who spend lots of time reflecting on things. For the thinking introvert, their introversion is driven by their minds typically turning inward, regardless of the social setting they find themselves in. You may see a thinking introvert staring off into the distance, not because they are bored or rude but because they’re contemplating something that has just popped into their head.

Thinking introverts have vivid imaginations and can spend hours working on creative projects and pondering over life’s great mysteries. They are drawn to anything that lets them take a deep dive into their mind.¹

It might be ADHD, it might be thinking introversion, or some cocktail of the two. But it now seems clear that the medication helps me stay more present with my family. I was quite happy to get that feedback from my wife, because now I can stop wondering whether that little pill is helping and just take it.

Come on, chin up. Sure, things might get worse by some measurements. Maybe much worse. But for all we know, that could lead to humans getting so fed up with modern tech that we embark on some sort of renaissance. Or it could be Skynet tanks rolling over skulls.

The point is: nobody knows. And wherever this is leading, at the moment, civilians like us can’t do a damn thing about it. This AI freight train is white hot and supersonic. Ain’t none of us slowing it down.

If we can’t see into the future, and we can’t control it, then what does worrying about it get us other than another gray hair?

I have concerns, too. I’ve got young kids, and I can’t help but worry a little about the world they’re inheriting. But I try to keep it that: worrying a little, and then reminding myself to let it go and live in the now.

And I call bullshit on the whole “web is dying” bit. The corporate web already sucks horrifically. And yet here we are, writing in our little indie blogs and following each others’ little RSS feeds. I fully expect we’ll keep doing this regardless the mountain of AI slop that piles up. If anything, that mess could make us double down.

Now, If you’re trying to make money and get Google-driven views on the web, I admit the situation is more complicated. That might get harder for a bit. But see above re: renaissance. Maybe this leads to something good, like new and better ways of paying for quality work from humans.

See also: Oliver Burkeman’s recent newsletter post, Let’s stop talking about A.I..

Time’s up. Bye. 👋

I clacked a few clacks and felt some solid Obama-not-bad vibes.

But what stuck with me most was one of the comments:

Ok, admittedly a great concept and super satisfying to play with.

However, god am I so burnt out on the Claude website aesthetic. It’s flattening the visual web experience. All the obvious tells like the “dot with tiny all caps labels”, and the one emphasized word in italics in headlines and the beige background.

A site like this would feel so much more… real?… if it had any sort of human design point-of-view. As is though, these sites all feel devoid of soul.

While my site doesn’t do those specific things (dot with tiny all-caps, etc.), it does feel flat. And it uses a vaguely similar color scheme to some of those Claude sites.

This has me wondering about giving my site a greater dose of human funk. Alas, as I’ve said a few times before, I’m bad and slow at web design, and therefore wary of burning time on it.

So for now, I’ll just keep wondering. Later, though, I may impulsively launch a foolhardy redesign. Stay frosty.

Today I’m revealing Altar II, an ultra-low profile (4.75mm) mechanical keyboard designed to replace your Magic Keyboard. It’s the keyboard I wish Apple made.

Altar II has fully mechanical switches, plus a magnetically detachable dial, haptic feedback, a red backlight, and a native macOS companion app.

As someone who likes mechanical keyboards but always ends up back on a Magic Keyboard, I was intrigued.

However, see the top HN comment:

To everyone who is about to pile on regarding the lack of a fingerprint sensor: you know it’s not possible, we all know it’s not possible, the creator of this project knows it’s not possible.

What are you adding to the conversation when you complain about it? So it’s a dealbreaker for you, don’t buy it.

I won’t comment on the lack of Touch ID, then.¹

What I’ll say instead: having a Mac mini as my desktop machine means I really like having Touch ID on my keyboard. To the point where it makes me wonder if I should have a Mac mini as my desktop machine. But that sends me into a fierce bout of Linux daydreaming, which I’m not up for at the moment.²

So this Mac mini is great I guess! And, cool as it looks, I won’t be buying that keyboard.

abbr --add jjc --set-cursor -- 'jj commit -m "%"'

This adds an abbreviation, which is a like an auto-expanding bash alias, that puts your cursor at the % character. So when I type jjc in my terminal and hit space, that expands to:

jj commit -m "<cursor>"

Handy!

My image in video calls looks so much better here compared to my home office, which is in our underground basement and has only one tiny little window to let light in. It’s amazing what natural light can do.

On that note, we’re starting a project soon to renovate our garage to do a number of things, including giving me an above-ground office with some nice windows. I’m looking forward to that. At least the part where the project is done. It’s going to be a slog to get there from here.

On the flip side, working here makes me appreciate my walking pad back home. My body, while bathed in wondrous sunlight, is feeling stiff and stagnant.

Here’s a pic of that mountain.

And one from the night before when the sun was going down.

No filters or processing because, as you can probably tell, I don’t know what I’m doing.

I had my ears checked this week by an audiologist, and my hearing is “borderline” normal except for a particular band of frequencies in which I have achieved “mild” hearing loss. And it just so happens that the typical adult female voice falls within those same range of frequencies.

My wife has some new leverage in our little debate.

The audiologist said it’s my choice whether to try hearing aids now, since my hearing loss is only mild, but that I’m probably going to need them in 5–7 years. Ever the curious cat, I decided to try a demo pair.

The effect was subtle at first. I even asked her, “Are they on?” I couldn’t tell. But I’m now three days in, and I’m appreciating them more and more.

We did family movie night this weekend, and it was literally the only time in our family’s history that my wife wanted to turn the volume up while I was completely fine with it. Usually I’m aspiring toward theater-esque volumes. I’ve been telling myself it’s because I want the IMAX experience at home, but perhaps I’m also trying to fill some gaps in the spectrum.

We went to a food cart for lunch yesterday and sat outside in a somewhat crowded area, with many people chatting nearby. My wife whispered at me, without leaning in, to turn around and look at a guy’s earring. (It was shaped like a huge tooth.) Instead of my typical response, “Huh?” I just turned around and looked. Her eyes lit up. “Ooh, this is fun!” She was referring to the fact that I can actually hear her whisper now. She whispers all the time as a means of covert communication around our kids. But it doesn’t work so well when I ask her to repeat her secret messages at volume.

I also suspect I’m feeling more alert, especially later in the day. Apparently this is common, as the aids reduce listening fatigue. It’s tiring when your brain has to put in extra work to make sense of sounds all day long.¹

So, the hearing aids may be here to stay. I have slight qualms about that, since I’m only 46. But they’re not too glaring a change in my visage. My wife, who sees everything, didn’t even notice that I had them in on that first day until I mentioned it several hours later. Maybe we can incorporate them into foreplay.

With that, I promise I’ll stop blogging about this blog for the rest of the month because I’ve been doing way too much of that lately. With one caveat: I’ll try to keep up with the Blaugust weekly themes, so I might bend that rule I just made if called for by the topic at hand.

On the x-axis, and just spiritually, I’m somewhere around the “Moved from Jekyll to Hugo people.” But I’m pretty high on the y-axis as well, which probably means I need to chill out.

Comic credit to Rakhim.

I found it last night after reading Brennan’s post, Free Speech and the IndieWeb, which discusses the reasons he doesn’t use Cloudflare. See also his longer post on this topic.

I have been feeling a little dirty about using Cloudflare for hosting this little indie blog. Plus, Cloudflare frustrated me this week when it wouldn’t update my broken site (my fault for breaking it; Cloudflare’s for not letting me fix it). So Brennan’s post provoked me to look for alternatives. I searched for options that allow deployment of static sites via rsync, and I soon beheld pico.sh.

I’d signed up and had my site deployed to pico pages in a matter of minutes. Truly, less than ten minutes, including the time to sign up and pay for a year of pico plus, which is only two bucks a month.

This is my deployment command:

rsync --delete -rv ./public/ pgs.sh:/blog

Et voila: https://sal-blog.pgs.sh/

What I’m liking so far

Other than reading docs, it seems I interact with pico.sh entirely over ssh and rsync. Auth is handled via ssh key. If I want to chat with the community, I use IRC with pico’s provided bouncer. Account notifications are provided via a unique RSS feed. This is all so geeky and awesome.

Deploys use actual rsync, so I can see which specific files are being updated, how much data is being transferred, etc. I was wishing for that kind of transparency when Cloudflare’s wrangler tool was refusing to update my site this week. This also gives me control over the deployment: I can upload only specific files or glob patterns, use --dry-run to preview the operation, etc.

Speaking to some of Brennan’s points, pico.sh has a Code of Content, which currently reads:

• Don’t upload content if someone else might own the copyright to it
• Don’t upload gore
• Don’t upload “hate speech” (i.e. demeaning race, gender, age, religious or sexual orientation, etc.)
• Don’t upload material that is threatening, harassing, defamatory, or that encourages violence or crime
• Don’t upload illegal content, nudity, or adult sexual content
• Don’t upload content to spam users
• Don’t upload content to employ social engineering tactics in an effort to trick users
• Don’t impersonate someone else

In short, don’t be an asshole. I like supporting a service that is willing to take a stand on that.

The founders are active on HN, and I like their vibe.

Their blog is filled with transparency about the evolution of the product, and they use RFCs with their community to help guide the direction.

There’s a lot to like so far, and I’m only an hour into this!

What’s next

I’ll keep poking around to learn more, run some tests, etc.

I need to get redirects working. It looks like they use the same format as Cloudflare. Wonderful.

If I’m still feeling this good about it in a day or so, I might just point my domain at it!

PT Mono was in the list of fixed-width fonts that made his cut. For some reason it stuck in my brain. Maybe because I hadn’t heard of it.

Later, I found myself perusing macOS’s Font Book app. There, in the relatively small Fixed Width collection of fonts, was none other than PT Mono. Turns out it’s a default font on macOS.

Coincidence, or obvious message from Zeus?

I set it as my VSCode font and have been quite liking it. It’s fun but not trying too hard.

And just writing this has sent me back down the font rabbit hole. This last paragraph is written using Ubuntu Mono. Why do I do this.