Peripheral Autonomy – A rambling preamble …

Where to begin? I haven’t written much software since I retired 2 years ago. I was inspired a few months ago to create a VSCode extension for generating music charts, which I had a lot of fun building. To be fair, GitHub CoPilot did most of the building after I corrected a few of its early errors. And amazingly enough, in a couple of weeks I was producing polished charts suitable for a band to play from (my band, but that’s another story).

That effort scratched a long time irritation I’ve had with music charts. But honestly I haven’t thought much about really doing anything serious with software since I retired. At least not until I read this article about Jesse Genet and her OpenClaw agent system. It was a fascinating read. The idea that someone set up a home system to do something for them was, well, inspiring. To me this is the promise of AI – not replacing my search engine but actually being helpful, taking on tasks that I would otherwise have to do and ultimately reduce my cognitive load.

Today, everyone use AI tools – ChatGPT, Claude, CoPilot, Perplexity, Gemini – and the list goes on. But mostly what we do with them is ask questions or feed them materials and wait for results. Admittedly, that’s mostly what I do as well. But the promise of AI is much more than that.

Businesses are starting to get the hang of it with AI call handling, customer support, and some back office functions. But I remember way back in 1984 when a small company called Apple released the computer for the rest of us – the Macintosh. When it first came out, with its black-and-white WYSIWYG screen, the thought was that these computers would live in our kitchens, pulling up recipes, family calendars and contacts, notes and reminders. But that wasn’t enough for the Mac to be successful. Instead someone developed FrameMaker, the first desktop publishing software, which essentially transformed the Mac from a true “home” computer into a somewhat specialized “business” computer. Today, that’s largely how Macs still sit – they target content creators (which is sort of a business application) and software engineers (who love the BSD base that MacOS is built on). PCs, on the other hand, largely followed the work at home track – they were used in business and people bought them for the homes so they could work at home.

There’s really a point here that I’m taking a long time to make – it’s going to be a long time until we get personal assistants that are really personal, not about business. There’s a good argument that maybe we don’t need it. But I decided to test that theory to see if I could build a system that actually improved my life by unburdening me.

I started setting up OpenClaw a little over a week ago and have made some quick progress. I’ve taken atypical approach to the setup because of my bias against running servers in my home. So for starters, my OpenClaw system lives on an Ubuntu VM in Azure. I’m also a cautious builder – I’m not quite ready to trust anything it does so I’m engineering it with safety and security in mind. I’m willing to spend the money on tokens up front but that won’t last forever – it ultimately has to be cost effective.

Most important, it has to do things for me that I would otherwise have to do. It’ll be a long haul project – I don’t simply want scheduled tasks that repeat themselves. I need things to be done without my intervention, except for perhaps approvals. My model is that of a chief of staff. A good chief of staff knows what it can handle on my behalf without approvals and what it can do with my approval. But more importantly, it can anticipate my needs and act to satisfy them.

Go big or go home, as they say. I’ve been fortunate in my first week of having a significant “Eureka!” moment that gave me hope that I wasn’t aiming too high. The next post will talk more about how I’ve approached this and what I’ve put together.

“Lead the way then, Mr. Netherton,” said Lowbeer.
Netherton did, imagining, as he climbed the stairs, a better world, one in which a relaxing drink would be waiting in the sitting room.
— from “The Peripheral” by William Gibson

Ironman 70.3 Hawaii (Honu) – May 2025 Race Report

Well that didn’t go as planned.

Screenshot

I read last year’s race report and thought I had written about it this year. This year, the weather was good and the swim wasn’t shortened. The winds at Hawi were headwinds and tailwinds, not cross-winds like last year. And it was hot, but it always is.

So the swim was pretty good for me. Somehow, I ended up being one of the last athletes in the water – I was lined up at the right estimated time, but people kept squeezing forward, so I got a late start. They clocked me at a 2:18/100 yd swim, which is respectable for me in the ocean. I saw a small turtle after the first turn and I managed to deal with the sunlight on the return.

And the bike was okay also. Last year I was 2 minutes faster but I attribute some of the lost time to slowing through aid stations – I was very thirsty after the swim and the Perpetuem (Chocolate) mixed with Kona Kola Nuun didn’t cut it, so I grabbed a water bottle at each of the aid stations, which made me feel better. My legs were a little slow out of the gate but after making the turn onto the Queen K, I felt pretty good. So not bad on the bike either.

The run. Well, I could just copy what I wrote last year. About 3/4 of a mile into the run, I was feeling bad. My left quad was very sore and my energy levels were down. I stopped running and started walking, eating an RXBar and a Honey Stinger gel, and then tried to get going again. That’s when I started cramping up – first the quad, then the hamstring and calf on my left leg. From there, it was walk and run. I really felt like quitting (first time I’ve ever felt like that), but I pushed on, hoping that I’d start feeling better. Eventually I got in a little running. I saw that my pace was around 14 minutes per mile, so started playing a game – when the pace slowed to 14 minutes, I ran as much as I could to get under 14 minutes. When I cramped up, I started walking until it hit 14 again. Up until the end, I managed the race this way. It wasn’t great, but I managed to run across the finish line with 7:44:18.

So what’s up with the cramping? I had plenty of electrolytes. Cramping is typically a neurological/electrical problem in the body. During race week, I don’t eat many nuts, which are usually salted, so perhaps my salt levels were lower for the race. But fatigue can be another cause of cramping. Maybe my training isn’t quite right, especially as I’m aging. I did a couple of bricks with 2 mile OTB runs after 60 miles on the bike and felt great, but on race day, that just didn’t carry over. Maybe I need longer OTB runs. I’m not sure yet, but this’ll be on my mind as I prepare for the Tri-Cities 70.3 in September.

If only I could age down next year instead of aging up!!!

Ironman 70.3 Hawaii (Honu) 2024 Race Report

Well, another half-Ironman is in the bag; that’s number 11 for me. I don’t think I imagined that I’d be doing these so regularly when I started in 2016.

This wasn’t my best race, by a long shot, but I don’t feel bad about it. These races are hard and if you don’t smash it, you learn from it and, perhaps, you learn a little about yourself in the process. The race was different than my other 70.3s because the swim got cut sort due to high winds (15+ mph). I thought that was actually an overreaction by the race officials, but they have their rules. So instead of the usual 1900m swim, we only got 800m. So things were weird from the start.

I had a very quick swim, clocking in at 2:11/100yds. Instead of wearing my Form goggles (which had to be replaced because the display failed), I was back to Roka R1s. I guess I didn’t have them adjusted properly because they fogged up shortly after the start and, when I pulled them away from my face (while swimming) to clear them, they ended up leaking for the rest of the swim. Perhaps I didn’t have the strap tight enough. I think in the future I’m going to carry a small water bottle to the start of the swim so that I can wet my face and goggles to create a better seal.

I had a great bike ride, really too great. It was very windy up at Hawi but I hammered out a 3:32:19. That was about what I always do this ride in, except for 2022, when it must not have been windy at all and I came in with a 3:20. The tubeless tires worked out great, I got through an entire bottle of Perpetuem and hydrated well with Nuun. It was extra windy this year, and I put in the extra effort to have a good ride.

I paid for that on the run, where everything that could go wrong did. I felt a little tapped out when I started and 3/4 of a mile in, I got a cramp in my left hamstring that took a while to work out. When I finally got going again, my back started really hurting. It wasn’t the usual disc issue that I have – instead the left side of my lower back was really sore. Maybe that was from the bike effort or maybe something else is going on. But the pain made it hard to run. I had Advil but unfortunately I left it in my run bag – now there’s a lesson for the next race. In the first couple of miles I also ended up with a cramp in my right hamstring.

Since it’s so hot, they always have ice at the run aid stations, so I figured out that I could ease the back pain by putting ice down the back of my jersey. Since I wear a run belt, the ice generally stayed in place. That, plus making sure I took the salt stick tablets got me running a bit more. There was a point where I felt pretty good and ran about 3 or so miles without walking.

I had to pause the running again when my stomach did a couple of flip-flops. I usually have some cola about halfway through the run for a bit of energy, but this time I started it early since I was already a bit sluggish after the bike. A little over halfway through the run, I was looking for a restroom. Of course I had passed one not too far back and as doing the math of backtracking versus running into the woods. Fortunately I made it to the next aid station and took care of the problem in a more civilized fashion. I was kind of a mess at that point, but got it together to run the last couple of miles to the finish line. So whatever time I had saved from the short swim was spent on the run. Ugh!

What will I do differently next time? First of all, I think I have to be done with cola drinks. I should just eat the Honey Stingers that I know work well in my stomach. I carried two of them, but never at them. I probably also need more than two. I usually eat a bar (like an RxBar) in T2 but I didn’t do that yesterday – not sure why I skipped it but that would have helped my energy levels. I probably needed to eat one on the bike in addition to the Perpetuem. I need to figure out why I’m getting the leg cramps and back pain. I drank nearly 2 bottles of Nuun on the bike ride (no cramping there) but somehow cramped up on the run. Maybe I need to double to Nuun dosage on the bike ride – two tablets per bottle instead of one. And I realized the Perpetuem doesn’t have significant electrolytes in it, so maybe that’s something else to fix. I think I also need to do more brick workouts and longer ones as well. Maybe run 5k off-the-bike more frequently.

These races are always a bit of a science experiment. After 14 races (11 70.3s and 3 full IMs), you’d think I’d have this dialed in. Things rarely go perfectly in these races and we’re all human and make mistakes. I have another 70.3 in September, so that’ll be another opportunity to do a little better.

I do want to give a huge shout-out to all my friends who were at this race. I had friends who just came out to see me on the course as well as a group of friends that volunteered at one of the run aid stations. In addition, I met one of the race referees the day before and she remembered me and said “hi” from the back of her motorcycle while I was doing the Kawaihae climb-out – that was super-cool and very motivating! All of the cheering makes a difference in addition to the camaraderie between the athletes while we’re on the course. I had several great conversations during the run as we helped keep each other moving forward.

Aloha!

Harvard Caps Harris Poll on Israel

The Harvard Caps Harris poll this month has an interesting set of data regarding support for Israel. In particular, the anti-Israel protests are not reflective of the overall sentiment of US voters. That’s not to say that there aren’t real consequences to the Jew hatred occurring in our streets and on college campuses or that brainwashing of young adults hasn’t happened, but these numbers show that a majority of voters are supportive of Israel. #StandWithIsrael #BringThemHomeNow

https://harvardharrispoll.com/wp-content/uploads/2024/04/HHP_Apr2024_KeyResults.pdf