C

AI and cybersecurity. Education, research, and good-faith debate. Evidence over tribalism.

[DHH @dhh]

Fascinating mind virus ripping through parts of open source right now. Listen, the code is good or bad whether it was written by math or by mind. Coder or clanker. Merge good code, reject bad code. You don't need a fucking code of conduct or contributor covenant for that!

[JD | RoyalCities @RoyalCities]

My favorite part of the game recomp discourse is being told by non-programmers that devs enjoy manually decompiling

How in the olden days, they would gaze lovingly into raw hex for 14 hours a day, carefully reconstructing functions by hand

Wait till they find out about Ghidra

[Kristof @CoastalFuturist]

How it feels to ask the Chinese ai models for help on something Claude won’t agree to

[C @nuancedresearch]

@zexyishere @tactmedic52 Sorry mw2/skate 3. There’s a Minecraft one too though

[C @nuancedresearch]

@zexyishere @tactmedic52 Oh yeah you’re right about that. I was talking about the mw2/minecraft cross over. That doesn’t use the decomp. The guy working on the decomp never released it

[C @nuancedresearch]

@zexyishere @tactmedic52 That’s the actual MW2 decomp. Different than what OP is referring to

[C @nuancedresearch]

@MichaelWaitze @Gavin_x1 Why is there this assumption that defenders are not using AI themselves? Claude security literally lets you use Mythos preview to evaluate your repo

[C @nuancedresearch]

@jandenhale 💪💪💪

[C @nuancedresearch]

Me

[sophie @netcapgirl]

there’s so much confusion about what’s happening in AI because if you don’t use coding models every day it’s hard to understand how quickly progress is accelerating but if you do, you likely have AI psychosis to a degree that it’s impossible to explain it in an understandable way

[C @nuancedresearch]

@SyntheticBeef Amazing! Thanks for sharing

[C @nuancedresearch]

For now. I follow this stuff closely and the direction the industry will go is more reliable models. We’re at a point where the level is intelligence is “good enough.” That will continue to improve, but the ability to task an agent with something and know that its output will be correct without human in the loop is the ultimate goal

There will always need to be someone that keeps at eye on the big picture, but the day to day will be completely automated

[C @nuancedresearch]

I’ve been working in the industry for over a decade. It’s standard practice now and some of the most brilliant engineers at FAANG now spend their days directing swarms of agents. The job has changed. Models like Astra and Fable are extremely competent at what they do

Google for example used an agent swarm to identify a bunch of areas in their infrastructure that needed optimization and made those changes

[C @nuancedresearch]

@Cmonitsmeguys @pogbum4040 @TheIshikawaRin Yeah let’s put down others! No one is allowed to have fun unless I agree with it!

Only me and those I approve of can make mods and write code 🤪

[Andreas Kling @awesomekling]

Stages of grief about the effect of AI on open source.

Denial:
- AI code is slop! Nobody serious will use this.
- Open source will keep working exactly like it always has.

Anger:
- These AI slop PRs are ruining GitHub!
- Ban AI-generated code!
- If you close PRs, you're not open source anymore!

Bargaining:
- Okay, AI can write code, but only if a human reviews every line.
- Maybe adversarial AI review will make arbitrary external code intake safe again.

Depression:
- Wait, if code generation is basically free, what is a contribution anymore?
- Is GitHub-style open source culture just going away?

Acceptance:
- Open source is still open source, even if the contribution model evolves.
- Good issues, test cases, benchmarks, specifications, and ideas become the scarce inputs.
- Upstream may stop accepting arbitrary code while forking becomes much easier.

Bonus stage! Reconstruction:
- Projects redesign around maintainer-controlled agents, issue-driven development, stronger test/validation infrastructure, and new ways for contributors to earn trust.

[C @nuancedresearch]

Do you know what big o notation is? A binary search tree? Hash collisions? Recursion? Stack vs heap? Amortized complexity? Graph traversal? Dynamic programming? Race conditions? Mutexes? Deadlocks? Memory management? Database normalization? Indexing?

No? Cool. I could hand you the exact same AI model and ask you to solve a real software problem and you’d have no idea what to ask it, how to structure the problem, or whether its answer was even correct

[C @nuancedresearch]

Or learn to make a game. Plan the whole game out on paper. Create your assets. Create a pseudocode diagram of the games architecture and systems. Feed that into AI and have it turn that into a project plan. Have the AI implement the first “slice” of the game. Maybe just the first level. Test it, communicate feedback to the ai, have it make improvements, test more, rinse and repeat. 500 prompts later you have an amazing game.

Just an example of a modern workflow. It’s a tool. You just have to figure out how to make it augmented your existing skills and workflow

[C @nuancedresearch]

@theojaffee Define conscious? We don’t even know the biological mechanism for it. Seems foolish to apply it to machines at this stage

[C @nuancedresearch]

@notblvdes @NightmareFoxxyy I don’t get how you can look at your responses and assume any sort of moral high ground. Just childish and loser behavior all around

[C @nuancedresearch]

@notblvdes @NightmareFoxxyy Exactly my point. You’d rather stroke your own ego than learn new things and new perspectives. That will take you far in life im sure… Luddites as they say

[C @nuancedresearch]

I’ve been in the industry for over a decade and have been writing code since childhood. Intellectually dishonest is not wanting to hear the perspectives of those more experienced than you by resorting to blocking and muting. It’s always some 20 something year old artist or newbie programmer trying to talk down to people like me. Linked thread is further evidence of my point

[LM @loyalmoses]

I was a programmer my entire life, worked extremely hard, very successful, my code made millions…

Last night, I spoke to Xcode and told it to create a Swift app for MacOS that had a drop zone for HTML files and to convert the animation into 4k still screenshots and assemble them into a ProRes 422 .MOV.

Kept editing in Final Cut Pro and 20 minutes later, 2400 lines of code were written and it works absolutely beautifully. It added common quality of life features without asking.

I do not recommend anyone start down the traditional programming path. AI is the disruptive technology that turns entire industries upside down. Learn accordingly.