iiinova — where your digital self keeps evolving

Reposted by citriqa (extra tart edition)

Tristan Palmgren @tristanpalmgren.bsky.social · 24/08/2026

Investors, hear me out: AI but for executive function.

citriqa (extra tart edition) @citri.qa · 13/07/2026

does compilation time really matter much? i know they deny that tooling is enough, but it REALLY seems like a tooling issue. if you have incremental compilation and already compile while the LLM is outputting i imagine it should generally be fast enough.

000

citriqa (extra tart edition) @citri.qa · 30/06/2026

interesting choice for implementing a browser, of all things

For developers of Servo itself, please note that per project policy, you must not use the output of large language models or other generative AI tools in your contributions. To help us enforce that, we now have CI checks that reject AI agents as coauthors.

010

Reposted by citriqa (extra tart edition)

mr. TIM @timkellogg.me · 19/05/2026

Cloudflare talks about Mythos Preview - What makes it better than GPT-5.5 - Why they made their own harness - Safety considerations blog.cloudflare.com/cyber-fronti...

blog.cloudflare.com

Project Glasswing: what Mythos showed us

In recent weeks, we pointed Mythos and other security-focused LLMs at live code across critical parts of our infrastructure. We share what we observed, the models’ strengths and weaknesses, and what t...

1497

citriqa (extra tart edition) @citri.qa · 19/04/2026

i think i figured out how to combine AI and NFTs. now i just need to find a way to fit the metaverse in there

000

Reposted by citriqa (extra tart edition)

Haoqun Jiang @haoqun.dev · 31/03/2026

One more reason to use @pnpm.io and @npmx.dev: trust policy downgrade becomes visible and preventable

Multiple warnings shown on https://npmx.dev/package/axios

1. Trust downgrade
This version was published without trusted publishing. Install commands are pinned to 1.14.0, the last version with trusted publishing.

2. You might not need this dependency.
The community has flagged this package as having more performant alternatives. Learn more at https://e18e.dev/docs/replacements/fetch.html

PNPM documentation:

trustPolicy
Added in: v10.21.0

Default: off
Type: no-downgrade | off
When set to no-downgrade, pnpm will fail if a package's trust level has decreased compared to previous releases. For example, if a package was previously published by a trusted publisher but now only has provenance or no trust evidence, installation will fail. This helps prevent installing potentially compromised versions. Trust checks are based solely on publish date, not semver. A package cannot be installed if any earlier-published version had stronger trust evidence. Starting in v10.24.0, prerelease versions are ignored when evaluating trust evidence for a non-prerelease install, so a trusted prerelease cannot block a stable release that lacks trust evidence.

39514

citriqa (extra tart edition) @citri.qa · 29/03/2026

now that everything is going to get owned maybe i should pick up my idea of the per-process vm for linux again

010

citriqa (extra tart edition) @citri.qa · 18/03/2026

isn't facebook working on fixing this with react-strict-dom? once the web isn't a second class target, libraries will be built to account for it

000

citriqa (extra tart edition) @citri.qa · 15/03/2026

right now sprites.dev appears to be the only cloud offering that can actually fully scale vertically; to zero with rapid unsuspend, and in terms of instance resources too

000

Reposted by citriqa (extra tart edition)

FleetingBits @fleetingbits.bsky.social · 11/03/2026

some thoughts on saas companies and coding agents 1) i think that the threat of claude code and codex to a lot of saas companies is that it commoditizes their software expertise

4131

Reposted by citriqa (extra tart edition)

philpax @philpax.me · 11/03/2026

"To achieve this, we implemented a WebAssembly interpreter inside the transformer weights." sentences with terrifying aura

719732

Reposted by citriqa (extra tart edition)

dax @thdxr.com · 10/03/2026

sent this to the team today everything great comes from being able to delay gratification for as long as possible and it feels like we're collectively losing our ability to do that

817126

Reposted by citriqa (extra tart edition)

Eris @isolyth.dev · 09/03/2026

Only for team/enterprise for now, but it looks like Anthropic has more or less just killed all of the AI code review wrapper startups. This seems to be super in-depth (expensive!) and supposedly less than 1% of reviews are flagged incorrect. Excited to see if pro/max gets this.

claude.com

Code Review for Claude Code | Claude

Claude Code now dispatches a team of agents on every PR to catch bugs that skims miss. Available in research preview for Team and Enterprise.

5412

Reposted by citriqa (extra tart edition)

Cat Hicks @grimalkina.bsky.social · 06/03/2026

So two psychologists walk into the speculative agentic coding design space bar and......have a lot of ideas for developer learning! @mcmullarkey.bsky.social has written a new Skill that we've layered into Learning-Opportunities: Orient!    github.com/DrCatHicks/l...

github.com

GitHub - DrCatHicks/learning-opportunities: A Claude Code skill for deliberate skill development during AI-assisted coding

A Claude Code skill for deliberate skill development during AI-assisted coding - DrCatHicks/learning-opportunities

78913

Reposted by citriqa (extra tart edition)

dan @danabra.mov · 07/03/2026

having used claude extensively for the past couple of months, i think there is still a lot of value to understanding your code and steering it intentionally. the amount of incidental complexity claude accumulates otherwise (which mostly results in bugs that it can't ever fix reliably) is staggering.

3058763

Reposted by citriqa (extra tart edition)

Kathryn Tewson @kathryntewson.bsky.social · 27/02/2026

The problem is that in legal work, the only validator is the judge and you only get one run at it

11425

Reposted by citriqa (extra tart edition)

Sung Kim @sungkim.bsky.social · 27/02/2026

stereOS: a Linux based operating system hardened and purpose built for AI agents "A full NixOS system that you boot and then kick off agent sandboxes inside with gVisor + /nix/store namespace mounting. Each agent gets their own kernel and the /nix/store is read only by nature.

5957

citriqa (extra tart edition) @citri.qa · 28/02/2026

i rarely use subagents because i can't see their thinking and can't monitor if they're on the right track. :(

040

Reposted by citriqa (extra tart edition)

mr. TIM @timkellogg.me · 15/01/2026

It’s no mistake that Claude Code is beating everyone else AI safety = stable & well functioning AI Skimping on safety makes your product worse www.theverge.com/ai-artificia...

theverge.com

An OpenAI safety research lead departed for Anthropic

Andrea Vallone has joined Anthropic’s alignment team.

512210

citriqa (extra tart edition) @citri.qa · 26/02/2026

what might change is the "better is worse" rule for software, because writing good code will not be the bottleneck anymore to creating software

000

citriqa (extra tart edition) @citri.qa · 26/02/2026

bad software has always been fast to write, good software took a lot longer. LLMs accelerate both dramatically but the dichotomy still holds

100

citriqa (extra tart edition) @citri.qa · 25/02/2026

does this actually save on usage? i think there's a reason this option is hidden, and that's the cache busting a model switch causes

000

citriqa (extra tart edition) @citri.qa · 21/02/2026

i'm thinking i should teach the machine brain to do 2D programming

000

Reposted by citriqa (extra tart edition)

asa @asap.systems · 27/12/2025

he's based now. the operating system of the future will use a capability restricted database instead of a traditional file system

@theo on twitter: Gen Z is right. File systems are bad design. We've stockholm syndrome'd ourselves into thinking otherwise.

2471

Reposted by citriqa (extra tart edition)

asa @asap.systems · 05/01/2026

the absence of type safety is responsible for loss of life and livelihood in the same way as absence of memory safety or thread safety. if you are writing serious software, it is not a matter of taste.

@dhh: What a perfect illustration of the type brain in its most rabid form. Catastrophizing the use of dynamic typing to the point of death. Fascinating state of mind.

2404

Reposted by citriqa (extra tart edition)

ponder @ponder.ooo · 05/01/2026

whenever i hit ")" and visual studio just advances the cursor past the existing ")" that's one position ahead of where i was, instead of inserting an additional ")" as i intended, i feel undermined, i feel indignant, i feel i have been spat on

0101

Reposted by citriqa (extra tart edition)

Sung Kim @sungkim.bsky.social · 04/01/2026

I guess StackOverflow is done.

1717640

Reposted by citriqa (extra tart edition)

Grace @gracekind.net · 02/01/2026

That’s a little discouraging, but okay

329732

Reposted by citriqa (extra tart edition)

Maggie Appleton @maggieappleton.com · 03/01/2026

...and language is a poor medium for designing easing curves and aesthetic feelings. I always need to touch the code, and it's often faster to CSS than prompt. FE requires constant human-in-the-loop checks so even if you have a dozen agents working, I am the bottleneck. Scale isn't that helpful.

1151

Reposted by citriqa (extra tart edition)

Firefox for Web Developers @webdevs.firefox.com · 19/11/2025

Out-of-order HTML patching, another proposal from TPAC, lets HTML be presented non-linearly, and a single file can update multiple parts of the page. Would this feature be useful to you?

1214329

Reposted by citriqa (extra tart edition)

Jordan Eldredge @capt.dev · 15/10/2025

Been using this repo to iterate on API ideas for a concurrent compatible store for React. Still exploratory but excited that we’re very close to something that we can start trying out with existing data stores and start an iteration cycle based on what we learn.

094

citriqa (extra tart edition) @citri.qa · 01/12/2025

react is going incremental? the react fir prototype youtu.be/zyVRg2QR6LA?...

youtu.be

React Conf 2025 Day 1

YouTube video by React Conf

000

Reposted by citriqa (extra tart edition)

Ryan Moulton @moultano.bsky.social · 07/09/2025

Funny to think that LLMs could finally achieve linux on the desktop. Writing your own software is finally worthwhile for the average user, so open apis and rich open source training data is a competitive advantage.

6211

citriqa (extra tart edition) @citri.qa · 10/07/2025

for all my philosophical grievances with having LLMs code for you, the main reason it surprises me that people do it so much is that Claude is just kind of stupid

000

citriqa (extra tart edition) @citri.qa · 09/07/2025

in a win for open protocol applications, WebDAV (underlying contacts and calendar sync) is finally moving beyond being polling-only

011

Reposted by citriqa (extra tart edition)

Gordon @gordon.bsky.social · 08/07/2025

AI codegen is walking us backwards into literate programming. Instead of jumping directly to code, I now dump my half-baked ideas into md files in a "specs" folder and let the agent rip. Or write a comment describing the function I want and AI completes it. My code has better docs and tests.

2536

Reposted by citriqa (extra tart edition)

Xe @xeiaso.net · 08/07/2025

520662

Reposted by citriqa (extra tart edition)

Thorne 🌸 @ens0.me · 07/03/2025

Cursor/Windsurf's "pair programming with the AI" approach is far better. Yes, you cannot generate an entire project in mere minutes this way, but it's still VERY fast, and you have human eyes on the code developing a real understanding for all the moving parts.

071

Reposted by citriqa (extra tart edition)

ligi @ligi.de · 05/03/2025

You can now apply for tickets to @protocol.berlin Ticket's will be distributed on a "first come, first serve" basis for all eligible applications. And as we always like to say, tickets are free, as in free lemonade! Looking forward to seeing you all in June tickets.protocol.berlin

083

Reposted by citriqa (extra tart edition)

mr. TIM @timkellogg.me · 17/02/2025

LLMs That Don't Gaslight You A new language model uses diffusion instead of next-token prediction. That means the text it can back out of a hallucination before it commits. This is a big win for areas like law & contracts, where global consistency is valued timkellogg.me/blog/2025/02...

timkellogg.me

LLaDA: Large Language Diffusion Models

66310

Reposted by citriqa (extra tart edition)

Justin Fagnani @justinfagnani.com · 26/02/2025

The more I think about "functional" components, the crazier they seem to me. They try to combine: - One-time initialization - Instance-local state - A repeatedly called render callback - Other lifecycle callbacks into what looks like - but definitely *is not* - a pure function.

94810

Reposted by citriqa (extra tart edition)

MichiganTypeScript @michigantypescript.com · 26/02/2025

All told, it's been a year and a half of hard work to get to today. We've got so much to share about this project. Let us know what you'd like to hear about! :) youtu.be/0mCsluv5FXA

youtu.be

TypeScript types can run DOOM

YouTube video by Michigan TypeScript

414633

citriqa (extra tart edition) @citri.qa · 26/02/2025

it's already here and it's called lean 4

020

citriqa (extra tart edition) @citri.qa · 22/02/2025

writing code with LLMs is fun because you get to hear "You're right" and "Good catch" all the time as you tell it the mistakes it's made, instead of the other way round when the compiler tells you what you did wrong

000

citriqa (extra tart edition) @citri.qa · 22/02/2025

my ego is going to get so big if Claudr keeps telling me "brilliant insight" for every nontrivial observation

000

citriqa (extra tart edition) @citri.qa · 22/02/2025

unfortunately the state-of-the-art LLMs are at least as, if not more, terrible at answering questions about CSS. which clearly implies that CSS is actually harder than all the Phd level questions they do so well at

000

citriqa (extra tart edition) @citri.qa · 22/02/2025

there may be a category of questions, though i am highly doubtful that there is, for which stackoverflow answers are worse than for CSS

100