iiinova — where your digital self keeps evolving

And I would also appreciate any development assistance, 'cause I'm just a vision-science enthusiast and I'm sure there's stuff I've missed (even apart from it currently being an alpha-release product with no warranty that it won't randomly crash).

010

Logan R. Kearsley @gliese1337.bsky.social · 04/10/2026

Right now, one has to build it themselves, but I hope to get pre-built Windows and Linux releases ready soon. (It should be Mac compatible, but somebody else will have to create the Mac release because I don't have a Mac.) If you are able to run it, I would appreciate user feedback.

110

Logan R. Kearsley @gliese1337.bsky.social · 04/10/2026

I just GitHubbed a new alpha-version #worldbuilding tool: github.com/gliese1337/X... It's a tool for modelling arbitrary visual systems (including aliens you make up!) and exploring how they respond to spectral stimuli, so you can compare visual perception across different species.

github.com

GitHub - gliese1337/Xenovision: Compare different visual systems against arbitrary stimuli.

Compare different visual systems against arbitrary stimuli. - gliese1337/Xenovision

130

Logan R. Kearsley @gliese1337.bsky.social · 03/10/2026

There is a thing called a "prediction market" which in theory can generate value while rewarding competence. They are sorta-complicated way of paying for expertise. The things currently heavily advertised as "prediction markets" are... not that, and just feed addiction.

010

Logan R. Kearsley @gliese1337.bsky.social · 03/10/2026

Now that I have caught up, I think season 3 can be summarized as "What If A Man Was Even Larger?"

000

Logan R. Kearsley @gliese1337.bsky.social · 01/10/2026

Also: vast overuse of "real" and "genuine".

130

Logan R. Kearsley @gliese1337.bsky.social · 26/09/2026

What should be the equivalent of the Iron Ring for software engineers?

120

Logan R. Kearsley @gliese1337.bsky.social · 26/09/2026

I wouldn't. It seems like a straightforward extension of the programmer-to-woodworker pipeline.

011

Logan R. Kearsley @gliese1337.bsky.social · 25/09/2026

Not all poster sessions are the same, but generally you will not be giving a talk as much as casually chatting with people filtering by. So, work on the elevator pitch, and having quick answers for whatever you think might be common questions.

120

Logan R. Kearsley @gliese1337.bsky.social · 21/09/2026

Speaking as a software engineer myself, it does allow doing some things faster... but it also makes a lot of mistakes, fails when there isn't any public documentation for the relevant domain, and leaves you with a codebase that you probably don't understand in depth and can't debug without it.

001

Logan R. Kearsley @gliese1337.bsky.social · 21/09/2026

Blue and green are spectrally adjacent, and covered by a single word in many languages, but I would not call them "close".

110

Logan R. Kearsley @gliese1337.bsky.social · 21/09/2026

Not a joke! There really are blue pumpkins. And red ones, and yellow ones, and white ones. The red ones tend to be really watery.

210

Logan R. Kearsley @gliese1337.bsky.social · 20/09/2026

Blue ones are the best for baking, in my experience.

100

Logan R. Kearsley @gliese1337.bsky.social · 17/09/2026

OK, so you don't actually understand dualism vs. materialism. Got it.

100

Logan R. Kearsley @gliese1337.bsky.social · 16/09/2026

Fun work quote: "We don't need to beat the enemy to everything having to do with AI; if they all kill themselves with AI, we can hold back on that."

040

Logan R. Kearsley @gliese1337.bsky.social · 16/09/2026

If you assume that replicating the physical substrate of some complex system *won't* result in a new system with the same behavior, *that* is implicitly dualist.

100

Logan R. Kearsley @gliese1337.bsky.social · 16/09/2026

That's a weird definition of dualism that neither seems very useful, nor like it would match most people's intuition. I am assuming that minds are encoded in the state of a physical substrate. A dualist would posit that there is a second, non-material thing.

200

Logan R. Kearsley @gliese1337.bsky.social · 16/09/2026

Who ever claimed that reducing a neuron yo a zero or one would provide sufficient detail? There's an astronomical gap between "the operation of human brains probably doesn't rely on every detail of its quantum state" and "we only need one bit per neuron". That's just ludicrous.

100

Logan R. Kearsley @gliese1337.bsky.social · 16/09/2026

Hence, it should be easy for a simulation to trick the senses.

000

Logan R. Kearsley @gliese1337.bsky.social · 16/09/2026

If you assume that a brain simulation will only produce acceptable results if you simulate the environment around the brain as well, then yeah, an acceptable environmental simulation would be one that successfully replicates sensory experience.

100

Logan R. Kearsley @gliese1337.bsky.social · 16/09/2026

What? How do you get *that* from what I wrote? Physical systems don't have to work like a computer to be simulated on a computer. And simulating a dead brain would just leave you with a simulation of a dead brain, not magically turn into a living one.

000

Logan R. Kearsley @gliese1337.bsky.social · 16/09/2026

Only if 1) you assume that you need to capture the entire quantum state, and 2) you want to do it non-destructively. But if the classical states are a good enough snapshot, or you're fine with move semantics instead of copy, then it's "just" engineering.

100

Logan R. Kearsley @gliese1337.bsky.social · 16/09/2026

Huh. I would've thought just the opposite; you can't upload a soul, but maybe you could record the state of all of your neurons and simulate their activity in a computer.

110

Logan R. Kearsley @gliese1337.bsky.social · 16/09/2026

Ah, well that would be embarrassing for him, then.

100

Logan R. Kearsley @gliese1337.bsky.social · 16/09/2026

For certain values of "like". It is, indeed, a non-dualist argument, but not one that I find at all *convincing*.

010

Logan R. Kearsley @gliese1337.bsky.social · 16/09/2026

Yes. So if that's the strongest non-dualist argument available, I see no reason to find Will's claim in this case "embarrassing".

110

Logan R. Kearsley @gliese1337.bsky.social · 16/09/2026

So the argument is just that no simulation at less than atomic resolution could be good enough? That seems implausible. You *definitely* don't need anything near atomic resolution to replicate the environment well enough to trick human senses.

340

Logan R. Kearsley @gliese1337.bsky.social · 16/09/2026

Is there any non-dualist argument that some process in the brain *can't* in principle be simulated on a computer?

140

Logan R. Kearsley @gliese1337.bsky.social · 13/09/2026

Any progress on this that I missed in the last year?

122

Logan R. Kearsley @gliese1337.bsky.social · 11/09/2026

(This was part of a discussion on the potential for speciation due to cultural divisions in bird song.)

010

Logan R. Kearsley @gliese1337.bsky.social · 11/09/2026

Today in "linguists are not kidding when they say that your capacity for language creativity enables you to understand sentences that have never been said before in the entirety of human history":

I don't know that. I don't doubt that some humans will get freaky with someone they can't talk to, though, and some bird girls might be the same.

120

Logan R. Kearsley @gliese1337.bsky.social · 11/09/2026

Yeah, I'd say "the second Friday of every month". And to get the other interpretation, "every second Friday" or "every other Friday", with no mention of months.

040

Logan R. Kearsley @gliese1337.bsky.social · 11/09/2026

Maybe? But one interpretation is sufficiently obvious that I have to really pay attention to make the other one work.

110

Logan R. Kearsley @gliese1337.bsky.social · 11/09/2026

How do we know they die? Maybe they're like Immortal Jellyfish!

031

Logan R. Kearsley @gliese1337.bsky.social · 11/09/2026

It works pretty well for large-format kids' books. But mostly for toy bins.

110