Gemini* and Gopher

• Personal/Opinions

  • Manifesto

    The following was written shortly after my arrest. I am currently groupless, having resigned from the Racketeers, so ignore the signoff... The Consience of a Hacker... by The Mentor 1/8/86 Another one got caught today, it's all over the papers. "Teenager Arrested in Computer Crime Scandal", "Hacker Arrested after Bank Tampering"...

  • Songwriting/Slow

    The sky is pink and blue (my brain suddenly remembers an old tweet, with a picture of such: "sky says trans rights"), and I've just come in with the dogs, a few minutes before 8 am, ready to attack my white whale. I am trying to finish a song. I have been trying to finish a song for a long time now, ever since getting back to guitar other than classical last year.

• Technology and Free Software

  • Meditation on LLM Interpretability

    I have been spending more time on interpretability lately, doing several more projects to see if I can find a new angle for understanding what LLMs are doing. I even brought out a few old friends, including HTM, to see if tricks and obscure old ML methods could make a dent in stopping humanity from going out... That's mostly a failure. Welp, expected: if interpretability is anywhere remotely close to hard, we'd already have done it. But that got me thinking about something that's been in the back of my mind: what exactly is the limit of interpretability?

    Seriously. What do we mean when we say we are doing interpretability? There's already a very concise and accurate explanation of how the model works: the model itself! It's not like LLMs are magic and we need to burn a part of our soul as an energy source. They are just numbers and math operations applied in a very specific, predefined order. RoPE here, MatMul there, attention somewhere else, etc. The real question interpretability is asking, from my understanding, is how to establish high-level understanding and control over the model. Note that I didn't say human understanding -- the model's weights may implement complex operations via a short chain of nonlinear vector operations, and it could take many more words in English than there are bits in the weights to explain them with all their important nuances. For LLMs or humans alike, exploding circuits and algorithms implemented in weights into a long English explanation doesn't help anyone, and there's no guarantee that the English explanation is even remotely understandable.

  • Roughened Optical Fibers Turn Soft Sheets of Silicone Into Three-Dimensional Shape Sensors

    Researchers from the Department of Mechanical Engineering at the Massachusetts Institute of Technology (MIT) have detailed a soft, conformable sheet capable of tracking shape and movement as it bends and twists — targeting wearables for healthcare and more.

    "In physical therapy, you often need to gather patient data on how they are moving and stretching their arm during rehabilitation," explains first author Qufan Yu of the team's work. "Surface shape sensing could be very useful in that way. Our setup is targeting fully soft and stretchable sensing surfaces so it’s potentially safer for interacting with a human or a soft environment."

  • OVERREGULATION

    I hurt my neck again a week ago, probably from spending too long peering down inside the rear passenger door of my Jag (#1) while fiddling with trying to fix the exterior door handle. Cue a few days stuck on the couch, followed by a week of decreasingly-light work that might have finally got too heavy assembling some things to sell yesterday, from the feel of it. At least I had quite a backlog of downloaded documentaries to work through, including another old economics series to follow The Age of Uncertainty and Free to Choose (2020-10-11.1The_Age_of_Uncertainty.txt): Commanding Heights (2002).

    I haven't actually watched the last hour or so yet, since I started getting better (healed by the power of economic deregulation?), so I won't attempt a review. In essence it's clearly biased towards the right, occasionally mentioning increased wealth inequality in countries that adopt free-market policies, but almost as an insignificant side-note. But it presents an interesting practical economic history, including in often-overlooked South America.

  • Internet/Gemini

    • No ATmosphere in Geminispace?

      I notice a pattern that people with capsules share “Mastodon” Fediverse handles, but not “Bluesky” ATmosphere IDs.

      I wonder if there is a cultural alignment, with Mastodon being much more “just us nerds” than Bluesky. Or maybe it’s just an artifact of many capsules having been started when it was the only viable replacement for Twitter.

    • October 9, 2026 at 12:00 PM

      You can now reply to my content via webmentions, Bluesky, and the Fediverse. That means that anywhere you see this, you should be able to respond and it will appear on my site!

    • Booklist of chrisod

      While "surfing" the Gopher sphere, I came across the Gopher burrow of chrisod [1].

      Among other stuff, there are two interesting lists on this burrow:

      - Best books chrisod had read in 2026

      - Best books of 2025

    • Cryptography of the Internet - Part 1

      Now that I have talked about and Private-key Cryptography (Part Base article) and about Public-key Cryptography (Part 0 article), I am writing about the way that GnuPG handles its so called Public-key Cryptography; which should be more accurately described as a Hybrid-type of Cryptography (between and Public-key and Private-key Cryptography). I call this Hybrid-kind of Cryptography as Session-key Cryptography.

    • Stuff; Fatigue, Typesetting and Gemrolls

      Apart from that, I've been typesetting for a little event in December. Had to do the very daunting task of contacting authors for permission, but 2 of 3 already replied positively. So, that's big positive excitment and relief. 🎉

      Spent yesterday afternoon battling Affinity for page numbers, though. Never again, shall I typeset a dwarsligger! [1] (Lies… I will again, because that format looks fun.) The challenge is to have page numbers that go up only every other page. So, numbered lists. BUT! for ease of assembling, I put numbers on the 'upper' page as well. And those did not want to cooperate. Both styles, identical (bar the 'name'), one behaving, the other doing weird shit. I didn't figure out what was wrong. So, I deleted that style, copied the one that magically worked and renamed the field that needed renaming. So, it works for now and I shan't touch it again.

    • Whither Gemrolls?

      I am consciously trying to bring my attention to Geminispace. The aggregators and this BBS are very helpful in helping me discover new things, but I most enjoy (and expect to sustain) my own aggregation of gemlogs from capsules I find interesting. Also, I will scour the backlogs of those.