Lately, I have been super busy, burying my head in the sand with many projects.
Unfortunately, my “TODO” list has exploded out of control, which means that I’m into the time-management space where I’m robbing Jesse James to pay off my gambling note with Doc Holliday, or something like that. I feel like I’m flailing to get one thing done, or another, and that flailing is making it less likely that I’ll get the first thing done. To add to that, I have my own “AI CRISIS”. Let me explain a bit what AI is doing and will do to those of you who have TODO list problems…
(My office is now full of these things, stacked hither and yon)
That’s a Jetson Orrin Nano, made by Nvidia. It’s a ridiculously sweet computer for an old system administrator like myself: 8bg, dual gig ethernet (In case I need a firewall) 5 USB ports, an Nvidia chip with a bunch of CUDA cores for a GPU or AI, and a generous number of pinouts for displays, relays, inputs, etc. That’s Nano1, There are 5 more on a shelf over here, and 2 other Mini computers in cases, that don’t have GPUs, and some old Hewlett Packard headless desktops that you can get on Ebay now for $50. [The Nanos should cost about $100, but the price has gone through the roof lately because: AI] What’s nuts, to me, is that they can stack 2 NVme memory drives, so I could put more storage on one of those than I’ve ever needed, plus some flavor of Linux of BSD, and I used to automate entire corporations with less CPU and storage handy.
What’s it for? Home automation, and some performance/reaction art projects (Project: CHATTERBOX) (Project: FAKENEWS) (Project: AUTOBIO) – intend to make some little cedar wood boxes with paper sides vaguely reminiscent of shoji, so they can be scattered in places to watch things and discuss this or that with people. That has been very stimulating, since it’s causing me to have ideas about system integration and connecting my projects, and each of those makes me jump in another direction. While I am doing the computer stuff in the evenings, I am still spending my days bladesmithing and doing other metal-work, it’s just that both are now vying for my time. Along with other idea stacks.
I guess I may as well discuss some personal-ish stuff that some of you may or may not know about: in 2020 I suffered a serious brain injury, apparently from a fall incurred that winter, in which I slammed my whole body, headfirst, onto some ice. Looks like I had a brain-bleed of a sublethal sort but lost a bunch of mid-brain and limbic system. I didn’t even notice anything was wrong once the headaches started, but by 2022 I noticed I was walking differently, talking differently, my legs were weak enough that I couldn’t stand up unless I was able to pull up on something, my dick didn’t function properly any more, and my bowels operated differently. Wow, each of those things you can work around but when you put them all together behind an MRI on which the radiologist writes “substantial brain atrophy” suddenly you’re in a forest, not dodging single trees. It takes a while to see a neurologist out here in Pennsylvania, so I had several months in which I thought I had early onset dementia and was going to wind up like my mom (who at that time was dying of her own dementia) and I drank a lot of alcohol to wash down a lot of drugs and time passed quickly. One of my initial cognitive symptoms was almost a complete loss of short-term memory. For someone like myself who reads and designs and plans, that is catastrophic. Imagine reading a book and you get to page 5 and have forgotten page 1. I don’t mean it’s hazy, I mean, it’s gone. Yes, just like the famous HM, who was a Baltimore/Hopkins resident when I was a kid. That is why the tone of this blog changed – for one thing, I had less patience for bullshit and for another, it was much much harder to write anything coherent (imagine re-reading your article for context each time you add a word, which – incidentally – is how a large language model writes text). My way of dealing with it was to haze things with alcohol and drugs, and bull through by continuing to do things. For example, eye/hand blade-grinding does not require short-term memory so I was fine with handcrafts unless they required measuring things and remembering measures or complex assemblies. But I kept at it and oddly it seems as though my brain has finally routed around the damage. I’m not back where I used to be, but I can carry on a conversation, design things subconsciuously, iterate, test, elaborate, etc., enough to be useful to myself. I think that forcing myself to try to learn Japanese (using Duolingo) has helped me remap all those things, because language learning is really just a whole bunch of short term memory being comitted to long term memory and embedded in speech processing. That has something to do with why I let my TODO list go crazy: I no longer sit around uselessly wondering what I did for breakfast.
Another reason I was not writing so much is because I’m noticing another loathsome split in attitudes, namely what might be described roughly as “pro-AI” and “anti-AI.” I’ve done my share of arguing that issue, like I did regarding arguing with the faithful, with MAGA, and with anti-vaxxers: sooner or later, it’s not worth it because you keep repeating yourself and you’d rather be charging off toward new horizons instead of staying back trying to talk around a bunch of gomers who still believe in supply-side economics, or something like that. Worse, I’ve seen that the anti-AI crowd have taken to using the same kind of advanced pile-on irrelevancy reasoning as anti-vaxxers, and once you see that, it’s hard to respect their position enough to engage with it. If google’s data center, with its multiple copies of the internet, or Netflix’ copies of all of hollywood plus television, or Amazon’s massive catalog and compute farms – and bitcoin – well, if that didn’t set one’s heart a’flutter about energy and water use, don’t talk to me about AI. (As I predicted, the power requirements there are dropping fast) But we’re still going to blast past +4C in global warming, so it doesn’t really matter, unless you have kids or people you think should survive. That’s not an option. [Minor digression, if you really want the scare scenario it’s painted by Peter Ziehan The End of The World is Just the Beginning [wc] he points out that when the petroleum-based economy collapses, some parts of the world will return to wood and coal] Anyway, nothing is going to be allowed to matter; just make sure you’re not making it worse.
Avoiding social media is a good strategy if you find your hatred of humanity is running over. There are some of the weirdest and stupidest conspiracies on Instagram, and they keep smugly trotting out their same thing over and over and ignoring the fact that they’ve been rebutted over and over. That is a characteristic of religion, conspiracy theory, anti-vax, etc. Scientists come up with objections to your objections, while the faithful keep repeating their opinions, louder and louder. I haven’t quite unplugged there, yet, since there are still some artists I enjoy, but, god, politics is depressing. I don’t think any of you want to hear what I think about that, beyond +4C is going to take care of the republican party and MAGA and their weird children. As I write this, Donald Trump is in China fawning over Putin and Xi – three horrible dictators congratulating eachother for setting the world on fire. We humans have to stop putting the worst of us in charge of important things, but, again, +4C will fix that. It won’t make it better, but it’ll fix it.
So let’s talk about something nice!
This morning, I had an idea for an application, and asked Claude whether it thought it would be doable.

That was how I used to do it back when I had a company of software engineers working for me. Then, some discussion and a few meetings and a decision whether it was worth it, or not, and maybe some scheduling and testing. Total time: months.
The way Claude works, it’s more like, well:

Total time: 4 minutes, most of which was me scoping the problem. Then followed another 10 minutes of back and forth, “it would be nice if…” and Claude responding with new versions of code.
Remember: when someone sneers that AI aren’t creative and that they are just statistically regurgitating CS101 textbooks, they are not worth arguing with. This was all taking place while I waited in my peridontist’s waiting room from 12:40 until my 1:00 appointment. The whole application was staged and done by 1:00 just in time for my teef cleaning.

The entire application is just an HTML file full of a bunch of something or other. I haven’t even looked at it; why should I?
If you wish to interact with it, it is here
So that is why the Orrin Nanos are all over my office: I have several socially relevant, interesting, education, problem-solving automation systems on my bench, in addition to some axes and clubs and razor-sharp fish knives from the problem-creating department. Life must be balanced. I recently completed a bronze age-style Scottish dirk, including proper home-cast bronze fittings, and a wrought iron blade with some hidden modern steel in it. It’s a beaut. Next up: I am looking at making some cast bronze steel-ware for eating foods with. When I am not doing that, I am napping, eating, or making systems.
I have also gotten back into making Tshirts again, because it’s super fun to now be able to more or less instantly screen AI designs that look great. I haven’t figured out a good way to get any of them out into the community, though I have mailed tons of them to various tech weirdos from my old circle of computer security people. I don’t want to just toss them into the crowd like rolls of paper towels.

Leave a Reply
You must be logged in to post a comment.