(Note: I’m going to use AI here because it’s the parlance of the times. I continue to be very frustrated by the terminology flattening caused by continuous use of the term AI – the difference between LLM, agent systems, generative video and image systems, audio processing, transcription, virtual manipulation and ink detection, and more is enormous, but we flatten everything into a single pair of characters: AI. This lack of nuance makes us weaker as a society, but for the purposes of this post, I’ll take the battle as lost.)

The famous mathematician and talented communicator Terence Tao has been leading the charge, or so it seems, on thinking through some of this. He has invited mathematicians to write guest posts on his blog, a good number of which I have read. I quite enjoyed “What should we tell our students”, especially the metaphor of the “polytope-of-ideas”. Math is moving fast, but watching experts in the field have serious conversations about the field has been fascinating.

Yesterday Dr. Tao posted slides to a lecture Math 2.0. He highlights the change from a history of math where proofs were scarce to one where proofs were abundant, the challenges of recognizing the boundaries of the field and how AI helps and how it doesn’t.

The thing that I want to note more than anything is his model of problem solving: The ability to produce a verified artifact in pure mathematics without involving humans bypasses human understanding, and reduces ultimately the ability of humans to produce good results in the world of applications.

Human endeavour is ultimately just that: Human. I suspect the inhumanity is why people react so poorly to AI art and why art seemingly has to show it is human now. The challenge faced by so many fields now is going to be redrawing the lines: Where augmentation is legitimately helpful and where it starts to sap the human endeavour. Different fields will draw these lines differently: It seems likely to me that painters probably won’t land on the side of ‘lots of augmentation’ – but other fields are going to have to walk this line.

Even the humanities will need to walk these lines. A recurring theme many fields are discovering is: “ChatBot: Solve My Problem” is rarely the thing – instead, many fields are discovering that these tools are augmentations for those with expertise, and horrible preemptive amputations for those who do not yet have expertise.

The bottleneck will soon become not research findings themselves, but the attention of experts in niche topics.

At the end of the day, each field needs to figure out: what exactly is the endeavour? Math is discovering (or reminding itself) that it’s not proofs. For other fields, the answers here will be diverse – and so too will be level of augmentation.

In my own field, the answers will be incredibly diverse! For some programmers the endeavour is the programming, the code is the art, and for them no level of AI augmentation will be acceptable. For others, sometimes the code isn’t art – it’s a means to an end, a disposable artifact to serve some other need. For those programmers, AI can be a helpful augmentation, reducing tedium and allowing one to focus on the actual task at hand.

In 2023 I wrote about Crossing an AI Rubicon. The world has only gotten stranger in every possible dimension since then. On the AI front, we’re all having to grapple with what it means for these strange tools to exist.

I just hope that more fields can find their Terence, and navigate this time with care and compassion.

Edited-To-Add: This was also a very good, skeptical read.

It’s nice to feel like I am not alone.

• I don’t run an auto shop

• I don’t provide services to RV owners

• I am not trying to grow my business

• I am not chasing leads

• I am not drowning in paperwork

I’m used to spam, but what I find so crazy about this spam is how incredibly narrow it is, like they have me badly confused for some other Matthew Gaudet, who really does run an auto shop. It’s also really effective at getting past the spam filtering my email provider uses.

Why couldn’t they have confused me for a chef, philosopher, or thoracic surgeon? I’d really love to see what their spam is like.

It’s a very good essay. What’s fascinating and horrifying too is the way in which this essay, written 10 years ago almost, has its finger on culture pulses that continue to echo today.

There’s so much of this that I would love to clip and quote, but I’ll leave this final one here:

An e-reader is a wonderful thing, so I found myself in the market for a new one. Shopping was hard, for reasons I was kind of surprised by:

• Kobo has eliminated black-and-white from it’s middle of the road e-reader lineup.

  If I wanted the most direct match to my lost reader, I’d be looking at the Libra Colour… but I didn’t want colour! It’s lovely in theory, but in practice it’s still not there, in my experience of looking at it in store. It’s noticeably lower contrast and at least a little less pleasant to read.

  As someone who is mostly looking to read books I really had a hard time telling myself that colour would be the right choice.

• I was quite interested in Boox: the idea of a full Android e-ink reader was appealing, as I would be able to read articles from my RSS feed easily, and growing skills around Android is always interesting to me. I know just enough adb to be dangerous.

  Of course, then I started to think more: The whole thing I like about e-readers is that I’m not “online”. The idea of being able to install arbitrary apps on my e-reader was probably more of a a downside not an upside. Also acquiring the B&W Boox Go 7 in Canada was just complicated enough that I hemmed and hawed on it for weeks.

One reader which stood out to me, and actually was the thing I pulled the trigger on first was the XTEINK X4. A colleague showed one off when I was on a work trip, and I was quite impressed. Given how affordable it was, I figured even it it was not to be my “main” e-reader, it would be a fun thing to explore, and apparently was quite hacakable. It turns out my timing on getting this was pretty good, as they’ve discontinued that model, though a replacement does seem to be coming.

While I was waiting for my XTEINK to arrive, I was dealing with the my lost Kobo, making sure whomever found my device couldn’t buy books, when I stumbled upon a refurbished Kobo Libra 2. This was exactly what I had been looking for – B&W Kobo, familiar form-factor. Even a small upgrade on a few dimensions. Perfect. I clicked order.

These two devices are quite different in a lot of ways, so I’m actually going to talk about the separately for now.

The Kobo Libra 2

The biggest reason I bought a Kobo, in addition to the X4, is that I like the ability to take e-books out of my local library. The integration of this is pretty darn good in a Kobo, and so I figured I’d like another Kobo.

Still, I was doing so much e-reader research, that I started coming across interesting things I’d never really seen before. For example, I learned there are open-source software options for Kobo that I’d never heard of before. The one I’ve dug most into is KOReader.

KOReader

KOreader is fascinating. It’s way more than a reader, but not quite a full firmware. When you open it, it takes over the whole Kobo, and can be used to control things like wifi, USB sharing, and more… but you can quit, and get back to the original Kobo interface.

One really appealing thing, knowing I was having the X4 en-route was that KOreader has a device-to-device progress sync, and there’s an implementation for the X4. So you can read on one device, upload your progress, then download it to another. It’s not quite the very-pleasant ambient sync capability that Kobos have, but it works on any e-pub you have, keyed by hash of the e-pub… so it works on Standard EBooks for example!

Let me get this right out of the way: KOReader doesn’t quite map right to my brain. I’ve been using it for months and I still find it a touch un-intuitive though it’s getting easier. Many people complain it’s confusing, and they’re not wrong. However, there’s some really wonderful stuff in there; For example: Kobo has this horrible problem when jumping from book to book of the text size jumping around horribly. Totally fixed in KOReader.

There’s a wonderful Calendar view of your reading which is just fascinating (though makes you notice the gaps when you’re reading things like gasp Paper Book)

There’s also some incredibly strange things in there… like a terminal emulator and an SSH server:

[root@kobo ~]# ls
bin         drivers     fs.md5sum   libexec     lost+found  proc        sbin        tmp         var
dev         etc         lib         linuxrc     mnt         root        sys         usr

I have a feeling I’ll be plumbing the depths of the KOReader for a good chunk of time.

The XTEINK X4

When I got it, I discovered right away that the default firmware is at best described as… not great. However, I had anticipated this and done my research, and heard great things about the alternative firmware: Crosspoint Reader. Flashing this was super simple: Plug in device and flash directly from Firefox.

It’s simple but effective. You load books over wifi or MicroSD card.

CrossPoint has quite a nice little WebUI when you’re connected via WiFi: You can change settings, log into your KOReader sync and more.

Reading

The experience of reading on the X4 is super nice. It’s shockingly light, very quick overall to read, and the screen is actually really good, though, I miss the front-light more than I expected – being in perfect light just happens too rarely in reality.

There’s a built in font downloader and a good selection of fonts. There’s an anti-aliasing story, but after spending some time experimenting I came down on the side of skipping antialiasing. Partially because it looks fine without, and partially because with it enabled it means page renders are two phase: no-AA render, followed a few hundred ms after by a AA rendering pass.

Image rendering is totally serviceable a bit slow.

One thing that has surprised me extremely is how central the weight is to the experience of the X4. It barely feels like it weighs anything, and it actually makes the Kobo feel quite like a brick! Really makes me wish more gadgets chased lightness!

Battery life is extremely good. The lack of frontlight and only turning on wi-fi when syncing means it’s a real sipper.

Hackable

It’s quite neat to see all the various things people have done with the X4. If I ever got bored of it or perhaps if the battery stops charging well, I could turn it into a little Trmnl. Honestly, it’s so neat I’ve wondered about acquiring one just for hacking. I’d love to find out if it’s bootable without a battery, is converting it to a status screen without the battery is just such a lovely idea.

Some Caveats

I would mention a few things I’ve noticed:

• At least on mine, the X4’s wifi modem is weak. By this, I mean that I can only connect basically when I’m in the same room as my router, where no other device cares so much.
• When considering firmware selection, it does seem like it’s possible to get into a cul-de-sac: be cautious when evaluating them.
• ePub rendering is very simple; which mostly means that sometimes more complicated formatting gets lost: block quotes, etc. I did also spot a few cases of ligatures hitting a fallback font.

Some Resources I Discovered on My Journey

It’s been neat looking into e-reading in 2026 with a more open mind than I have previously. I found a few neat resources to share:

• Nico Verbruggen has done a really wonderful job of creating some beautiful fonts for Kobo devices. Give them a try!
• For the X4, there’s a third-party app which is quite nice for using as a website clipper, called “Send to X4” which works pretty good for purpose. It makes it nice to use it to read really long articles
• The KOReader Compendium is alternative manual that I occasionally understand better

Conclusion

E-reading in 2026 is still pretty great, and I’m excited to see new generations of devices. Actually, Crosspoint has a whole bunch of devices that are supported now, including some very neat looking options like the LILYGO T5.

I’ve jumped the below video to the demo: Wow.

What a very fun show with a very charismatic performer.

I really liked this post from Duane Brett which reminds us… this is mostly an illusion.

Unfortunately this doesn’t mean we get to ignore the election. Albertans must stand up and push back on this governmentally-bolstered false separatism. The only plausible way that this tinderbox, that Danielle Smith chose to light, is extinguished is if we forcefully reject the question — and even then I worry that it may be too late, and that we will be dealing with this bullshit for decades.

Here’s a few links:

• No on the nine

• Forever Canadian

Of course, not being a real smoker, I coughed. I didn’t particularly like the smoking part… so I sucked back on it as fast as I could.

As I started to feel the effects, I felt great. The world was sharp, and I felt like I could cut through problems. I suddenly understood why smoking had taken over the world for a while.

Immediately afterwards I threw the rest of the pack in the garbage, knowing full well that if I ever smoked a second cigarette, it would be the beginning of addiction and dependency. That day built a huge amount of empathy for smokers, as I knew full well that I could be there too but for a few different choices.

Anyhow, this is a post about why I haven’t posted on Mastodon in a while.

Facebook was great when I first got on there. For young people seeking connection, it felt like suddenly we’d gone from having a small web of people to a huge one. Of course, it didn’t last. At least to me it seems like the social graph got bigger, but then we started to realize that posting everything to everyone wasn’t so smart, and then it all fell apart as we tried to curate and sort who could see what.

Twitter was great when I first got on there. Funny people, smart people, influential people. You felt like you were watching a whole river of incredible conversations – and you could just… partake. Sometimes you’d even get a reply! A hero read your tweet! Of course, it didn’t last. I was complaining about it feeling toxic in 2018.

Mastodon was great when I first got on there. A small community of interesting people who had self-selected into this world. The linear timeline felt like the features you wanted, federation would be the salvation from degradation of commercial interests.

Unfortunately, what I have realized more than anything else is… I’m just not built for social media. I can’t seem to figure out a way where I partake and I’m happy with it.

There are people who can be “social smokers”. I myself am a “social drinker”. At the same time, nicotine and alcohol addicts exist; the human experience runs the gamut.

For me, I think I will choose to just abstain from social media. It’s better for me.

I’ve basically not logged into Mastodon in months except to post a couple of my blog posts. I realized at some point that every time I opened the page I was rolling the dice: If I was lucky, I’d find something cool, get a nice dopamine hit. If I wasn’t lucky, I’d find someone who was mad. Or something would be posted that made me mad, or sad, or scared.

When I stopped logging in… I just felt better. Again.

Hmm. Maybe I’ll really learn the lesson this time.

I hope so.

I won’t post this there – different people feel differently and that’s fine, but I don’t need to lob some sort of meta argument onto Mastodon.

I may post other blog posts there. I do like people reading my writing sometimes!

What about blogs? What about newsletters?

I think they’re more my speed.

It’s something I absolutely can and should do more of. Recently I sent a note to a small musician when I realized their song had been popping into my head for nearly 15 years. I appreciate so many people in the world, but I don’t always say it.

I had been procrastinating on setting a similar letter to Victor Mair, who I know from Language Log, a blog I read and loved in the Google Reader era. What prompts this blog post is that, while debugging some feed reader issues I’d been having with Language Log, I realize… I am too late. Dr. Mair passed away on the 28th of June.

So, dearest gentle reader, I exhort you: Go forth and tell people you are grateful for what they put into the world, and do it eagerly: You never know when it could be too late.

But in 2026 — Fuck Energy. I’m so fucking tired of the fucking oil and gas business driving the terminology bus. This rhetorical spin of “Energy” and “Energy Superpower” is a fucking con-job, and I am so done.

Energy from oil and gas means burning shit. It means putting carbon in the air, or spending more energy to sequester the carbon as you burn it. It means dirtying up your air! It means living a shittier life, spending your (personal or societal) wealth pumping up the bank accounts of oil producing nations.

I THINK HUMAN THRIVING IS CONNECTED TO OUR ABILITY TO HAVE ENERGY BUT I AM ALSO AGAINST OIL AND GAS.

HUH. ISN’T THAT A NUANCED TAKE.

Technology Connections had a banger of a video talking about this in January (pre-Hormuz Cluster Fuck), and the point stands. You spend money on oil, you use it, and it’s gone. You spend money on Solar, Wind, Geothermal… and it’s there for the lifetime of that asset.

Why do we accept this rhetorical con-job? OH RIGHT OIL PRODUCERS HAVE MONEY AND A VESTED INTEREST IN CONFUSING THE ISSUE. Honestly, I’m really fucking disappointed in the Canadian Government for normalizing this bullshit.

Here’s a thing: Most countries on earth are oil importers. Most countries on earth are getting an object lesson in what it means to be beholden to world markets.

If I take any solace in the absolute bullshit of our present, it’s that the lessons of Hormuz are going to make a whole bunch of places realize that they should be looking elsewhere. I live in an oil producing area, and I have to say… methinks there’s a fucking train called electrotech coming for us, and when our province is fucked over by this… well, anyone who says “who ever could have seen this coming” should be laughed out of a room.

The world is going to be a better place when oil and gas are dethroned. I can’t fucking wait.

If you still haven’t encountered Angine de Poitrine, do yourself a favour and take a few minutes to watch the first 3 minutes of this. If you’re intrigued, just let me tell you it just keeps getting better.

This video was a fascinating watch. The thing I loved most about it is the way it felt like an artifact from a parallel universe, one where instead of the LED we got micro-fluidics.

The Tree

“A painting of single poplar tree in fall with leaves falling, lit just before golden hour, that evokes feelings of nostalgia and warmth.”

The Farmer “A fine art painting of a farmer holding a microprocessor"

Tuscan Hills

“A John Constable Painting of the Starship Enterprise over the hills of tuscany”

Van Gough Clouds

“Vincent Van Gogh paints Mammatus clouds over stonegenge” (spelling mistakes copied from previous prompt)

Ansel Adams City

“ansel adams photo of the moon rising over a city skyline, wide angle long distance shot high contrast dramatic” — once upon a time prompting was really a whole deal

Streamliner Modern

“A Streamliner Modern apartment building, line drawing, sketch”

Haiku Clip Art

“A black and white pencil drawing of a computer with arms using a fountain pen to write a haiku, simple art style suitable for printing, clip art”

Space Parents

“Two parents on a space station watching their child go to sleep on a baby monitor, digital art”

Cool Hijab

“A marble bust of a woman with a hijab and fashionable sunglasses, studio lit, soft focus”

Rockwell Bat

“An annoyed bat eating ice cream by Norman Rockwell”

Impressionist Baby

“an impressionist painting of a baby drinking from a bottle 1887"

Vaporwave Roman

There’s a whole bunch of things which factor into my interest in bouldering:

• There’s a ton of variety and novelty. One of the challenges I had with doing strength training at a gym was… well, a deadlift is a deadlift is a deadlift. And while you get a lot out of gaining strength through traditional training techniques, I really started to get bored, and getting bored meant I didn’t want to go, which meant… I stopped going.

  Whereas at my climbing gym, there are new sets every week. There are a variety of problem types in bouldering (slab, underhang, vertical), and each climb tends to be unique in its own way.

  Frequent new problems and frequent change means that coming to the gym is different every time.

• Speaking of problems, I really appreciate the extent to which climbing matches thinking and doing. While sure, if you’re strong, there are many climbing problems that can be done with brute force, more often than not there’s a much ‘easier’ way to climb a specific problem, and some problems really can’t be brute-forced without figuring out the ‘trick’. So there’s a huge aspect of climbing which is problem solving, which I really appreciate.

• It’s also a fairly full body sport. There’s not a lot of muscle groups that get missed by climbing, and can be pretty great cardio, and so climbing has been over the winter my dominant form of exercise.

I recently really had a bit of a revelation about bouldering that I wanted to write down.

In bouldering, problems are “graded”. There are some standard grading scales, though at the gyms I go to it’s basically just 1-9. For a long while I was climbing 3s, frequently being able to do them my first try (‘flashing’ them – climbing is filled with terminiology), yet aside from a single 4 accomplished in November, I’d not done any since.

Recently though I went from having done a single grade 4, to having done 6+ of them. It all started to come at once. The thing that’s interesting to me is how much of this was entirely a psychological barrier. There’s something fascinating about how merely having done ‘some’ of these routes suddenly made accomplishing more of them much easier. It is as if my brain and body suddenly gave me ‘permission’, in a way that hadn’t happened previously.

One of the most interesting videos about climbing I have watched actually covers this directly. Louis Parkinson of the Catalyst Climbing channel talks about mindset frequently (actually, I think it’s noticable if you watch his older videos how he’s actually grown on this front), and this is a whole video where he walks through the importance of mindset and how it is possible to train a more helpful mindset.

The revelation I really had is “grades are nothing but a number”, but that actually holds true everywhere.

• A faithful adaption of a well loved book. Don’t want to see a movie? read the book

• More excellent comedy work from Ryan Gosling

• Good Sci Fi results might beget more good sci fi movies? We can hope.

First and foremost I should say: It took me at least two and a half weeks of puttering around before I was remotely happy. There’s a few things that I did right, and a few things that aren’t yet supported which would have sped that up, but it’s not a “wow, you’re instantly capable” kind of addition. However, if you’re the kind of person who bought a Voyager, I suspect you already are OK with spending some time figuring things out.

Using a Trackball

I’ve long been trackball curious, but I’ve never made the switch. I am incredibly happy with Apple’s touchpads and their acceleration curves, but I have never gotten an amazing setup with my trackpad and my Voyager. Lots of motion, awkward hand positioning from having the trackpad between my Voyager halves. its’ particurly an issue when it comes to having more tent on the Voyager – it starts to feel like the trackpad is in a hole.

Some people have really nice clever solutions here, and honestly I may look more into them, but I really thought the trackball looked good.

Using the Navigator has shown me that there’s a bit of a mixed bag here. On the one hand, the the ball is quite nice. Haptically I really like it, and I like the easier access and fixed position relative to a Voyager half.

However, dialing in sensitivity has been a journey: If the sensitivity is too low, it feels like it takes forever to get your cursor anywhere, but if you set it too high, a trackball is very sensitive to your hand’s natural motion and hitting a target becomes a challenges of its own! At my preferred setting right now, things feel fine, yet I do feel like it’s still a bit too much jumpiness (there’s an aspect which feels almost like there’s too coarse a motion quantum too, but it’s purely a feeling)

Another interesting problem is momentum. I think Apple’s momentum based scrolling is remarkable. I had hoped that I would have physical momentum on my side with the trackball, but it turns out that the ball isn’t heavy enough or there’s too much friction, and so the ball doesn’t really show enough momentum compared to what I’m used to.

A Trackball With a Mouse For keys

Maybe I don’t brain good, but it’s been a month now, and I still am regularly getting mode confusion and problems. I’ve managed to fix quite a bit of the worst of it by banishing some of the most galling buttons to never-never-land, but it’s far from perfect.

Let me give you an example: The default mouse layout puts scroll lock on T – but given the default dwell time, I kept hitting CMD-T while the mouse layer was active, setting scroll lock and then not actually opening a new tab like intended. So shorten the dwell time right? Except, then, a moment’s hesitation when trying to do a click with f means you’re filling a channel with mistaken f strokes. Hilariously frustrating when you’re in a shared doc with colleagues as they watch you drop f in random paragraphs.

So I turned off auto deactivate. Now I need to click to deactivate. But this still means that if I jostle the ball or move my mouse for some other reason, I have to remember to hit a key to get the keyboard back into ‘being a keyboard mode’. But it’s fine if you hit most keys, because they’re transparent down to layer 0… except unimportant keys… like d,f and g. Oops. I moved scroll lock to r, so r is also no good.

To know if the mouse layer is active I set a layer colour, which helps. However, you can’t set a layer colour for scroll lock which would be helpful when you move your mouse and seemingly nothing happens, because you’re not in a scrollable container, but you’ve accidentally turned on scroll lock.

So you’re left in this situation where you’re never 100% confident what each button or what the ball will do.

Am I sure that this isn’t just me being dumb? I think this might just be the challenge of wanting to get too much into the same device.

Physical layout

I opened complaining a bit about my trackpad feeling like it was down a well. But the default mount for the Navigator has the opposite feeling – the ball is so high, you have to reach for it. I ended up getting my 3D printing guy to print me the official thumb shell. Strongly prefer this layout actually, though I miss my higher tent, and I keep thinking maybe I should get him to print me this with more tent.

Lots of Complaining…. you getting rid of it?

No, I think I’m going to keep it for now. When it’s working well it’s actually super lovely! I just think I need to keep tweaking some of this to get even more comfortable with it.

We’ll see what the update looks like in 8 months.

Here’s a perfect worked example of why they are important. The Economist writes about how college students can’t read. They talk about a study where students have trouble reading Dickens.

Well. Except it turns out that’s not the whole story! Mark Lieberman at Language Log did the work to find the study, which tells a more subtle narrative than the Economist. Read his post, it’s good.

This brings me to my main point: Journalism needs to learn to love the hyperlink. You want to develop trust in media? Link your sources. Think about a generation raised on media literacy classes that, where they are able to visit the source links for articles and compare; a generation that starts to understand the torquing and nuance loss that comes from a 6 paragraph opinion column. Think about the trust engendered by discovering that some outlets are really good at accurately reflecting their sources, and think about how the lack of links can become a signifier of dishonesty.

“But the sources journalists are talking about are complicated!” Sure! Not everyone will read them. Yet some will. Some will strive, some will learn, and grow. We have to give people the opportunity.

Maybe, just maybe, the Hyperlink is a key part of making sure that we understand the society we live in.

I love me a story with

• Characters who talk through problems. In fact, my least favourite narrative tropes are where the problems are created by two people, ostensibly who care for eachother, but cannot have even the simplest lowest-level honest conversation. See also “Could Have Avoided This!” Plot and Poor Communication Kills.
• Community Building, and community raising; the cynical part of me feels like this might be the most fantastical part of the book, but how I love it.

and of course, the requisite easy entrepreneurship.

It’s not a perfect book, but was a good distraction, and I’ve put the next on my borrowing queue at the library already.

It’s a book I actually wish was about five times thicker: you can feel the hours of interview that -didn’t- make the pages. I would have loved to hear more from many of the women she talked to.

It’s a hopeful book, though I did learn about Visitors traumatic sexual assault via this book, which I had never known about previously; a content warning of sorts.

As someone who was a child in the 90s, Visitor’s stories of Hollywood of the 90s are totally believable and yet horrifying. One of the through lines of the book is the way in which you sometimes cannot see the wrongs within which you swim. I have a few inklings about what we’ll say about 2025, but I wonder what we will see with more clarity in 2055.