The Art Of Not Asking Why

Finished reading: Rendezvous with Rama by Arthur C 📚

Ah, so this is hard sci-fi, a “category of science fiction characterized by concern for scientific accuracy and logic.” I watched The Martian without reading the book (but I did read Project Hail Mary), and I know this is another popular one I’ll add to my list.

Rama (for short) can be read as a textbook in a science class, and I wouldn’t be surprised if it’s used this way somewhere: how will a spaceship that sustains humans look like, and what makes it work? If you find yourself intrigued, this is the book for you.

It goes beyond that, giving us an interesting view of politics and society on a solar system scale, which I enjoyed (if you like that kind of sci-fi, I recommend “The Moon Is a Harsh Mistress”). A nice surprise is the existence of polyamory, or more specifically polygamy, which is the practice of marrying more than one partner (this also exists in The Moon Is a Harsh Mistress, making me curious to see what other sci-fi books have this). There are also a few common threads with The Expanse, including some themes I see in certain video games (which, I’ll have you know, can deliver a story better than movies and books).

This is my third library sci-fi book (which means it comes with a deadline, which I made in time) and, so far, my favorite. I’m taking a bit of a break from books at the moment, though I have a feeling it will be a short one.

At this point, three app-related updates in the last couple of weeks that I think are here to stay:

1. Arc Browser
2. Lillihub (for Micro.blog)
3. Kagi Search

Arc Browser

Available only for macOS at this point, this one is the most complicated, and I’m still getting used to it. Complicated is not fair of a work, let’s say, different. There’s much to say about Arc Browser and how it does things. Many of its features exist in other browsers or in well-known extensions; others are freshly new. Some of the big highlights for me are:

• Vertical tabs: This just makes sense on your screen once you get used to it
• Bookmarks/tabs/pinned tabs are essentially the same, the function change on the location in your sidebar. The presentations of those in your workspace is different, helpful, and hard to describe.
• Boosts: for those familiar with Grease Monkey, this is similar. It allows you to inject CSS and even JavaScript into your bookmarks, so the next time you visit them, they are customized to how you want to use them. Very powerful and very useful.
• Workspaces & Profiles: when you switch a profile (say personal vs. work), it is like opening a whole new browser in the same one. Different looks, different extensions, different search engine, different profile signed - all available with one “swoosh” of your finger or a keyboard shortcut.

There’s much more to this browser, and I barely scratched the surface. Download it and test it, but do it for a couple of days. It takes some time to get used to.

Lillihub

This is a beta client for Micro.blog, developed by Loura. It’s already my default Micro.blog client. Some of its immediately useful features include:

• When replying, checkboxes for people in a conversation in Micro.blog. Finally you can choose who to reply to in a thread!
• Drag and drop for picture posts
• Renovated Discover section, with a helpful left bar of all the emojis (categories here on Micro.blog), and buttons to Announcements, Challenges, and News.
• Pinned posts (cached locally) so you can always find that one post you want to get back to
• Helpful conversation tab, so you can focus on active posts with replies to participate in a discussion.

There’s more, and Loura might still be accepting testers. It’s still rough-looking visually, which is what you’d expect from a beta product, but everything works and works well.

Kagi Search

I just started using this Search Engine two days ago, and I’m quickly learning to like it. It just displays information better (better than Duck Duck Go, in my opinion). It draws information from Wikipedia, Reddit, YouTube, and more places you’re likely to click on without logging in. It has filters to a level I’ve seen only in Google so far (for example, the image filter allows you to filter by license, image type, HD/SD, Aspect ratio, etc.)

There are no ads. No sponsored links. That’s because you’re expected to pay for it ($10 a month if you hit the 300 searches limit). At first I thought, “What am I nuts? Pay for search?” But the more I thought about it, the more I realized how often I use search and how it makes sense to pay for a quality service like this if I don’t want ads or Google up my…

I think this one’s quickly becoming a safe bet, unless I find out they somehow do use users' data to sell to data brokers. That wouldn’t surprise me, since the opportunity is just so rewarding and so easy.but I guess we’ll see.

What about you? Thoughts? Suggestions?

Took a walk in the woods for some peace and quiet. I came across a circle made of branches.

Inside, another ring and a cross made of rocks with white and orange flowers.

I don’t know who or why, but there it was.

I was playing around with the new macOS Sonoma aerial screensavers.

I noticed there doesn’t seem to be an official way to remove downloaded screensavers (about 500MB MOV files each, give or take). The only way is to go to the system folder where they’re stored and delete them manually.

I guess it’s time to dust off the wiki and wire these things down 📖

Last week, I got the Sony WH-1000XM5 Wireless Noise-Canceling Headset 🎧. Nat works at a headset store, so deciding where to go was easy. I got to try on various headsets with two kinds of music I listen to often: Electronic and Classical. The Sony’s won by a large margin.

I still can’t get over the noise cancellation, which I now call “world canceling.” How good are they? Well, after a night out last weekend, I took the subway from New York Penn Station, a very busy transportation hub, back home. The “world cancellation” was so good that I couldn’t hear trains most of the time, only when they slowed down near the station. People moved their lips, but no voice came out. A homeless person gave me a jump scare by waving his hand in front of my face (I was looking at the tracks and didn’t see him). This was a bit unsettling, so I turned around to be more alert, using my eyes.

The headset is phenomenal. It cancels out firetrucks, yelling kids on a break from the nearby elementary school, and Obnoxious construction noises. It comes with an app that seamlessly allows switching from my iPhone to the Mac, depending on which device I’m using. I sure don’t miss having a messy long wire one bit. I’ve been spoiled, and I’m not going back 🎶.

Before I got my new MacBook, one of my worries was having it plugged to power 24/7. I do most of my work on it from home, and it’s plugged into a KVM switch to my external Monitor, Keyboard, and Mouse, along with my Windows PC.

Today, I noticed the above window for the first time (it shows up after you click the battery and charger icon in the top menu), and it was odd to see my battery is only at 80%. Alarmed, I checked Apple’s documentation about this. It turns out not only there’s no issue with leaving the battery connected to power 24/7 (per research I did previously), but macOS expects it from my habits and allows the battery to discharge to a certain level to keep it healthy.

Good morning. The sun is finally out, after being out of sight for the last… I don’t know, four days? I actually got excited yesterday when it came out just in time for sunset in the evening. And to think I’m trying to get into Cyberpunk Red now, a game that is all about gray, depressing vibes…

Sipping my coffee in the kitchen at my usual spot (see day in the life) I was thinking about how much work I have to do, the usual sense of not doing enough.

It doesn’t matter how much I worked yesterday and to what extent (I was drained toward the evening); the self-judgment is still there. The guilt. Feeling like I’m a phony. If I dwell enough on these thoughts, they’d borrow in with flashes of the past.

I’m learning to make peace with these thoughts and try to let them be. There’s no satisfying them, so instead, I try to give them a nod of acknowledgment and do my best to ignore them for the rest of the day. Some days it works better than others. I think that being open about those thoughts might help: if I put these emotions out there openly, I can’t be a liar. I’m not really hiding anything.

That’s enough for now. I think I’ll get myself to go out for a run, get it out of the way, let the skin soak some sun, and take a refreshing shower afterward.

Happy Wednesday!

Anyone can write instructions, they say, especially in the age of AI. Just write a list of steps and add a screenshot or two, and that’s it. I wanted to play devil’s advocate, so I fired up ChatGPT to see if - through the eyes of someone who writes and edits knowledge articles daily - these claims hold any weight.

Let’s start with a TL; DR I hope is rather obvious: Things are seldom black and white. Knowledge articles are not just set of instructions; they are also an authority behind a procedure, a responsibility behind claims, and grounds for support services when the instructions fail. AI instructions don’t contain these qualities.

The Scenario: Taking Photos with an iPhone

I wanted to use a standard, well-documented procedure, so I asked ChatGPT how to take photos on an iPhone. A quick look glimpse showed me ChatGPT’s instructions weren’t bad. It’s easy to get by if you’re familiar with an iPhone and basic camera functions, and most people probably are. Here are the instructions, pasted here for convenience:

How to Take a Photo with an iPhone

1. Unlock Your iPhone: If your iPhone is locked, press the Sleep/Wake button (located on the right side of most models) or use Face ID or Touch ID to unlock it.
2. Access the Camera App: Locate the Camera app icon on your iPhone’s home screen; it looks like a camera. Tap the Camera app icon to open it.
3. Choose a Camera Mode: The Camera app usually opens in Photo mode, which is the default mode for taking standard pictures. Swipe left or right on the screen to select different modes, such as Portrait, Video, Slo-Mo, Time-Lapse, and more, depending on your iPhone model and software version.
4. Frame Your Shot: Use your iPhone’s display as your viewfinder to compose your shot. Tap on the screen where you want to set the focus and adjust the exposure by sliding your finger up or down on the screen. Make sure your subject is well-framed.
5. Capture the Photo:
   • For iPhones with a physical Home button (older models), press the circular button at the bottom center of the device to take the photo.
   • For iPhones without a Home button (iPhone X and later), tap the white or red circular button at the bottom center of the screen to capture the photo.
6. Review Your Photo: After taking the photo, swipe left to view the picture in the Photos app or tap on the preview thumbnail in the lower-left corner of the Camera app to see it.
7. Edit and Share (Optional): If you wish to make edits to your photo, tap the “Edit” button when viewing the photo in the Photos app. After editing, tap the share icon (a square with an upward-pointing arrow) to share the photo with others via messaging, email, social media, or other apps.
8. Save Your Photo: Your photo will automatically be saved to the Photos app on your iPhone. It will be organized into your Camera Roll or other albums based on your settings.

Remember that your iPhone camera may have additional features like HDR, Live Photos, and various filters, which you can explore within the Camera app to enhance your photography experience. The exact steps may vary slightly depending on your iPhone model and the version of iOS or iPadOS you’re using.

Things Fall Apart

This doesn’t look too bad, right? As an iPhone user, I’d say so myself. The problems start to show if you don’t have an experience with an iPhone or taking photos. But let me take a step back first and give you the bigger picture.

You see, as a technical writer, writing instructions is usually the last thing I do. The first two important steps (that missing here) are research and testing.

I won’t claim to test everything I ever wrote instructions about. I often rely on other credible sources. In the case above, I could reach out to our Apple sysadmin for review and a crew of service desk technicians for testing. Since I have an iPhone, I would also test the procedure myself, grab screenshots, and take note of problems I encounter.

See that last paragraph in the instructions? It leaves it to the user to figure out things on their own. It has no room here. These are instructions for a procedure, not a workshop with creativity-exploring exercises. For a big organization, this could mean a day in hell for Desktop Support technicians who would need to help an overwhelming number of users determine their iOS version or “explore” the “various filters” within the Camera app. What filters? Are they needed, and What for? And these are just a couple of examples.

For research, I will first determine which iOS version the instructions are for specifically and highlight the “Apple terms” that might not be familiar to some users, like Touch ID or Face ID, or even the Home bottom (on Android, users had three buttons for a very long time). This leads to a good point: style guides.

At work, I try to follow the Microsoft style guide where possible (this is because this guide is one of the most common and available). A term like “Touch ID” would be in bold, indicating this is a specific term to iOS, not a common phrase like file or “camera” or “photos” (not to be confused with Photos, the app, which should be bolded). It’s also good practice to link references to these terms directly so users can catch on quickly.

Speaking of users, who are we writing these instructions for? Is it for Android users? iPhone users? Photographers maybe? When in doubt, I try to keep instructions at their most basic level without over-explaining terms that should be familiar - and that “should” is already a dangerous assumption. For better delivery, this article could refer users who need more information to Apple’s support article about the camera app and the iPhone manual.

And yet another issue while we’re speaking of the audience: level of access. Certain data (images included) should only be accessible to employees, sometimes specific employees with specific access. While this can be regulated by AI, deciding what level of access an article needs can be difficult.

In the example above, the instructions need to specify if certain pictures cannot be shared outside of the organization. There should be a link or a reference to what procedure should be followed (if at all) to share pictures. Note that while these instructions are not about sharing photos, which is a different topic altogether, we must assume that users who need to take photos might also need to share them. This kind of connection we make as humans is hard for AI to do, yet it may be critical.

These are just a few of the problems I see in the above article. It’s worth mentioning that the information needed to resolve these issues is almost never available to me by default at work. I have to find the right people who have the responsibility for the information and ask questions. It’s not uncommon that these questions alone can stall publishing instructions while the different parties involved debate the issues and try to come up with answers. Asking dynamic questions is a crucial part of technical writing.

AI is a tool, not a complete solution

At this point, I see AI as a tool to help streamline parts of a complex process. I believe many organizations (especially those involved in technology directly) rush to use AI as the new “cool” tech. That’s unfortunate because it hurts the quality of their services and products.

AI doesn’t cater to those at the margins with different needs. It is programmed to work with a “most probable use case,” not those exceptional times when something goes wrong. It’s in these times, when help is needed most, that AI fails.

With AI replacing humans, the need for innovation and improvement disappears because those depend on failure outside of the scope of automatic programming. Things go wrong when something unexpected happens, and the solution is usually an improvement to the system. With AI, responsibility disappears, and with it, morality and business ethics. We humans are not perfect boxes that fit into algorithms, no matter how much the CEOs of big tech companies want us to be.

I plan to keep using some of the AI tools available, but just as tools. I enjoy writing and learning about technology, and I can’t imagine letting AI take all the fun. Sure, it takes more time, but I’d argue it brings much more quality to the table.

Any idea how I would go at trying to get my data (check-ins) exported from Foursquare? There are a few scripts here and there that involve API, but it’s over my head…