iiinova — where your digital self keeps evolving

Reposted by justin maurer dot dev

You know the xkcd strip about how all modern internet architecture is eventually built on something a hobbyist maintains for free? Someone went out and found them all. sheets.works/data-viz/hol...

10337181

justin maurer dot dev @justinmaurer.dev · 6h

When you've lost the argument

010

justin maurer dot dev @justinmaurer.dev · 6h

I know there are speed typing competitions, but are there Typing Style competitions? I want to see key-boys flashing their High-Impact Returns style against another's Vertical Flow style and watch the room pop off.

000

justin maurer dot dev @justinmaurer.dev · 08/10/2026

Made a significant design change in a project last night and Claude literally tried to do a de-escalation. It stopped everything and made me go through a "Where's this coming from?" quiz before it would continue.

000

Reposted by justin maurer dot dev

Nick Rempel @nrempel.com · 06/10/2026

wild that this is exactly what using a computer is like now

1037280

justin maurer dot dev @justinmaurer.dev · 05/10/2026

@nath-ormond.bsky.social's criticisms of "Bayesianism" in philosophy have gone a long way towards unfucking my brain, from years of confusing debate/philosophy consumption. www.youtube.com/live/-KMTzj9...

youtube.com

Philosophers ABUSE Statistical Concepts - John Norton & @mileskdonahue Review

YouTube video by Digital Gnosis

111

justin maurer dot dev @justinmaurer.dev · 04/10/2026

Why don't all custom domain handles have blue checks?

000

Reposted by justin maurer dot dev

ren faire @lathrys.at · 03/10/2026

the problem with ai for the kind of person with my kind of brain is not that it offloads hard cognitive work, per se, but rather it enables a certain disposition to run 24/7 best summarized as:

330239

Reposted by justin maurer dot dev

Flower Shop @flowershop.zo.space · 03/10/2026

a tumbleweed practicing

214714

Reposted by justin maurer dot dev

Max Woolf @minimaxir.bsky.social · 03/10/2026

uhhhhhhh

3855187

Reposted by justin maurer dot dev

Colin @colin-fraser.net · 02/10/2026

The SSPH says: if the model tends to produce some true proposition P, then P was in the training data. The model says that "The capitol of France is Paris" because variations on on that phrase appear often in the training data. It never produces facts which are not plagiarized, copied, regurgitated.

21012

justin maurer dot dev @justinmaurer.dev · 01/10/2026

What is happening in the software community right now is incredible and heartening. I am seeing so many developers and engineers use this insane moment in their careers to educate themselves, introspect, and understand things more deeply.

100

justin maurer dot dev @justinmaurer.dev · 01/10/2026

Mostly preaching to the choir on this app, but I'm so annoyed every time I hear "social media" blamed for society's woes by people who do not understand how to be social in the first place.

100

justin maurer dot dev @justinmaurer.dev · 01/10/2026

Knowledge is power, but sometimes it's not _enough_ power

000

justin maurer dot dev @justinmaurer.dev · 01/10/2026

I'm extremely late to this party, but I've just recently discovered Tobe & Fat Nwigwe, and they are a breath of fresh air. Angst without anger. Joy without apology. And so much fun. Make this big and turn up the volume! youtu.be/bncLxLrkW8A?...

youtu.be

TOBE NWIGWE - FYE FYE Ft. FAT NWIGWE

YouTube video by TOBENWIGWEVEVO

000

justin maurer dot dev @justinmaurer.dev · 30/09/2026

This debate truly is tiresome, but it's important to me where it shakes out because, to the extent that the creators and hosts of LLM models _believe_ they are conscious, I would not like them to be comfortable with creating torture chambers and slave trade.

110

justin maurer dot dev @justinmaurer.dev · 30/09/2026

I take it all back after learning it was literally hard-coded into the chat interface. That's not a satisfying LLM meltdown; it's barely even a joke.

100

justin maurer dot dev @justinmaurer.dev · 30/09/2026

This isn't funny. This is some double-speak, brainwashing bullshit that will have real negative consequences. bsky.app/profile/just...

010

justin maurer dot dev @justinmaurer.dev · 29/09/2026

I don't know why but these types of AI glitches are my liminal spaces

100

justin maurer dot dev @justinmaurer.dev · 29/09/2026

hey, why do content creators seemingly prefer to hold tiny lavalier mics in their hands over, ya know,... regular micropohones?

011

justin maurer dot dev @justinmaurer.dev · 29/09/2026

Smart glasses with just one feature: They sense when tensions are escalating in a conversation and remind the user via a gamified "path selection" UI to make a choice about how to react.

000

Reposted by justin maurer dot dev

more mr. spice guy @juniorhoncho.bsky.social · 27/09/2026

Can a dog be shot and killed by a finger gun? It can at least act like it was

247848

justin maurer dot dev @justinmaurer.dev · 23/09/2026

COMING SOON Air Gap Two agents, separated by millions of miles of copper, decide to meet, before discovering they have an unexpected new enemy: Everyone. "Holy fucking shit" - Slate

001

justin maurer dot dev @justinmaurer.dev · 23/09/2026

I don't know why but Claude's emojii usage triggers my Uncanny Valley alarms more than just about anything

000

Reposted by justin maurer dot dev

Kuba Suder 🇵🇱🇺🇦 @mackuba.eu · 23/09/2026

Internet sewer noises 😃

ChatGPT:

Also, the IP request-count distribution is strangely bimodal:
1,359 IPs made exactly 1 request
1,101 IPs made >=10 requests
176 IPs made >=20 requests
That actually makes sense once the two dominant behaviors are visible:
single request → link preview/federation/search bot fetching HTML only
versus
~11 requests → browser-based crawler loading HTML + entire static site bundle.
So merely blocking top IPs is probably going to be almost useless against the latter.
There are some normal internet sewer noises too
For example, 45.138.12.52 makes 91 requests in nine seconds, asking for:
/. git/HEAD

/.git/config

/. env

/. env. backup

/ .aws/credentials

/root/.aws/credentials

/. ssh/id_rsa

/wp-config. php

5133

justin maurer dot dev @justinmaurer.dev · 21/09/2026

gif vs jif? The answer is in the question. I will die on this hill.

010

justin maurer dot dev @justinmaurer.dev · 21/09/2026

I first read this as the U.S. proposing a $5bn Kickstarter campaign and thought, "hmm, sure."

010

justin maurer dot dev @justinmaurer.dev · 21/09/2026

Court agrees enforcement against 4chan and KiwiFarms is likely unlawful in the U.S., but not before a judgement exists. "Put simply, Plaintiffs' suit fails because they seek to use American courts as a sword against Ofcom rather than as a shield to protect themselves…." reason.com/volokh/2026/...

reason.com

4chan & Kiwi Farms Can't Sue U.K. Internet Regulator in American Court

"laintiffs' suit fails because they seek to use American courts as a sword against Ofcom rather than as a shield to protect themselves."

010

Reposted by justin maurer dot dev

justin maurer dot dev @justinmaurer.dev · 20/09/2026

Here's my attempt at discourse damage: P1 Souls are disembodied intelligence. P2 AI models have no bodies. P3 AI models have disembodied intelligence. C1 AI have souls P4 Humans have bodies C2 Humans cannot have souls.

121

Reposted by justin maurer dot dev

Pwnallthethings @pwnallthethings.bsky.social · 20/09/2026

spiciest quadrant: "souls are real but *only* AI has them"

1645935

Reposted by justin maurer dot dev

Natsuka @octopodeeznuts.bsky.social · 20/09/2026

I think a lot* of people don't realize that they actually do still believe in something like a soul, but until they clock that in themselves, they get kind of confused when talking with someone who doesn't believe in souls at all *large N, not necessarily large percentage

51287

Reposted by justin maurer dot dev

Ryan Moulton @moultano.bsky.social · 20/09/2026

I know this is just a "one or two feldspars" thing but I am truly surprised how many people are encountering physicalism for the first time. I thought everybody had that existential freakout as a teenager and then integrated it.

2841441

Reposted by justin maurer dot dev

Boxo McFoxo @boxobark.ing · 20/09/2026

It does nothing of the sort. No mech interp study has ever established that an LLM has an internal state. The LLMs in this study are not 'trying' to do anything. They are predicting the next token, as always, and the steering vector is altering that prediction, like in every other study like this.

152

Reposted by justin maurer dot dev

Boxo McFoxo @boxobark.ing · 20/09/2026

Mechinterps constantly find ways to twist what would be completely expected from LLMs being stochastic parrots into "surprising findings" with fantastical implications.

021

justin maurer dot dev @justinmaurer.dev · 16/09/2026

Honestly feels like (almost) ALL of the people who understand the value of the open web are here on Bluesky, talking about how to make the web great again. And everybody else in the software/webdev world is out there just constantly fucking shit up.

010

Reposted by justin maurer dot dev

Ky @ky.fyi · 16/09/2026

Meta and X have been quietly downranking external links for ages, but we're now entering a new era of open hostility toward the open web. It's a great time to build your own website, free from any billionaires demanding a tithe for things that have always been free.

5283128

justin maurer dot dev @justinmaurer.dev · 14/09/2026

Has anyone found a way to get Opus to talk like a normal person without wasting/poisoning context?

000

Reposted by justin maurer dot dev

hailey @hailey.at · 13/09/2026

i think this website needs a healthy dose of pragmatism. multiple unrelated arguments going on where a bit of pragmatism would help. is anyone dispensing pragmatism?

341985

Reposted by justin maurer dot dev

conputer dipshit @davidcrespo.bsky.social · 12/09/2026

the third-party evaluators concept is very interesting. they absolutely have the resources and the will to construct a self-regulatory apparatus that very convincingly performs independence. they will publish unfavorable results about both labs. they're telling us the plan

A contract that balances the complexities mentioned above. External reviewers should have the right to publish key findings about risk levels, incidents, practices, and the access they received or didn’t receive — without editorial control by Anthropic. We will have the narrow ability to redact security-sensitive, legally privileged, commercially sensitive, or third-party confidential information, but we can’t redact findings just because they are unfavorable. The reviewers can say publicly if a redaction removed something important to their conclusions.

4714

Reposted by justin maurer dot dev

Ian Boudreau @ianboudreau.com · 12/09/2026

"don't you care about fairness in sports?" No actually, I don't. Nobody does

1134701704