Hacker News Reader

gloosxon AI Police Reports: Year in Review

>Create a test for intelligence that we can pass better than AI

Easy? The best LLMs score 40% on Butter-Bench [1], while the mean human score is 95%. LLMs struggled the most with multi-step spatial planning and social understanding.

[1] https://arxiv.org/pdf/2510.21860v1

gloosxon AI Police Reports: Year in Review

It's like saying google search is smarter than everyone, amount of information indexed by it has no human counterpart, such a silly take...

gloosxon Please just try HTMX

This fucker didn't even bother showing us some real POST requests. And he actually wrote JavaScript to make those work.

Why would you mock it on the client-side if HTMX makes it so simple??

gloosxon UK House of Lords attempting to ban use of VPNs by anyone under 16

Education?

gloosxon Australia begins enforcing world-first teen social media ban

"It's just kind of pointless, we're just going to create new ways to get on these platforms, so what's the point," said 14-year-old Claire Ni.

Claire Ni concluded it best. They are just going to find new ways. Imagine a kid stopping using something because of the law or government ban. Those lawmakers are just delusional if they think they can pass a law and the kids will stop using social media.

gloosxon Show HN: Gemini Pro 3 imagines the HN front page 10 years from now

of course LLM doesn't experience or feel anything. To experience or feel something requires a subject, and LLM is just a tool, thing, an object.

It's just a statistical machine which excels at unrolling coherent sentences but it doesnt "know" what the words mean in a human-like, experienced sense. It just mimics human language patterns prioritising producing plausible-sounding, statistically likely text over factual truth, which is apparently enough to fool someone into believing it is a sentient being or something

gloosxon I wasted years of my life in crypto

Reducing real human struggles to a punchline is exactly the kind of cynical detachment you can afford only if you have never lived through a failed banking system. If you had lived through what people in some countries deal with, you would not be making snarky comments

The reality is that crypto became a lifeline in places where the traditional financial system collapsed or simply abandoned people: Venezuela, Argentina, Lebanon, Nigeria are good examples of people dealing with real crises, using whatever tools actually work, including crypto currencies.

gloosxon Has the cost of building software dropped 90%?

From the creators of the 10x engineer:

0.1x engineering cost

gloosxon The "confident idiot" problem: Why AI needs hard rules, not vibe checks

>We are trying to fix probability with more probability. That is a losing game.

>We need to re-introduce Determinism into the stack.

>If it fails lets inject more prompts but call it "rules" and run the magic box again

Bravo.

gloosxon I wasted years of my life in crypto

>Outside of buying sex and drugs the only uses for cryptocoins are, and always has been, ransoms, scams and gambling

That is a very shallow take. There are real non-criminal uses for crypto that people in stable, wealthy countries often overlook. Millions rely on it simply to move money between family members across borders when traditional banking is slow, blocked, or outright inaccessible due to politics. In several countries, people use crypto to buy food, medicine, or basic goods because their local currency is collapsing or their banking system is dysfunctional, or their entire nation has been cut off from the global financial system as a decision of few politics persons.

Its fine to criticise the scams and speculation — there is plenty of that — but pretending thats the only use case ignores the people who depend on it for everyday survival.

gloosxon IBM CEO says there is 'no way' spending on AI data centers will pay off

It shows, quite an interesting wrapper over GPT with unauthorized access to prompting it you assembled there ;) Very much liked the part where it makes 1000 requests pulling 1000 comments from the firebase to the client and then shoots them back to GPT via supabase

take care

gloosxon RCE Vulnerability in React and Next.js

Of course you do, in certain cases making less round-trips to the server is just straight more efficient

gloosxon IBM CEO says there is 'no way' spending on AI data centers will pay off

It is always "I'm producing 300 projects in a nanosecond" but it's almost never about sharing or actually deploying these ;)

gloosxon OpenAI declares 'code red' as Google catches up in AI race

Try to ask it to write some GLSL shaders. Just describe what you want to see and then try to run the shaders it outputs. It can output a UV-map or the simple gradient right, but when it comes to shaders a bit more complex it most of the time will not compile or run properly, sometimes mix GLSL versions, sometimes just straight make up things which don't work or output what you want.

gloosxon It’s been a very hard year

If LLMs truly "understood architecture" in the engineering sense, they would not hallucinate, contradict themselves, or miss edge cases that even a mid-level engineer catches instinctively.

They are powerful tools but they are not engineers.

And its not about legal responsibility at all. Developers dont go to jail for mistakes, but they are responsible within the engineering hierarchy. A pilot is not legally liable for Boeing's corporate decisions, and the plane can mostly fly on the autopilot, but you still need a pilot in the cockpit.

AI cannot replace the human whose technical judgment is required to supervise, validate, and interpret AI-generated systems.

gloosxon It’s been a very hard year

A PM or QA can sign off only on process or outcome quality. They cannot replace the person who actually understands the architecture and the implications of technical decisions. Responsibility is about being able to judge whether the system is correct, safe, maintainable, and aligned with real-world constraints.

If AI becomes powerful enough to generate entire systems, the person supervising and validating those systems is, functionally, a developer — because they must understand the technical details well enough to take responsibility for them.

Titles can shift, but the role dont disappear. Someone with deep technical judgment will still be required to translate intent into implementation and to sign off on the risks. You can call that person "developer", "AI engineer" or something else, but the core responsibility remains technical. PMs and QA do not fill that gap.

gloosxon It’s been a very hard year

Even if someday we get AI that can generalize well, the need for a person who actually develops things using AI is not going anywhere. The thing with AI is that you cannot make it responsible, there will still be a human in the loop who is responsible for conveying ideas to the AI and controlling its results, and that person will be the developer. Senior developers are not hired just because they are smart or can write code or build systems, they are also hired to share the load of responsibility.

Someone with a name, an employment contract, and accountability is needed to sign off on decisions. Tools can be infinitely smart, but they cannot be responsible, so AI will shift how developers work, not whether they are needed.

gloosxon Program-of-Thought Prompting Outperforms Chain-of-Thought by 15% (2022)

The abstract actually says 12%

gloosxon It’s been a very hard year

It's not as simple as putting all programmers into one category. There can be oversupply of web developers but at the same time undersupply of COBOL developers. If you are a very good developer, you will always be in demand.

gloosxon Migrating the main Zig repository from GitHub to Codeberg

If unit tests are boring chores for you, or 100% coverage is somehow a goal in itself, then your understanding of quality software development is quite lacking overall. Tests are specifications: they define behavior, set boundaries, and keep the inevitable growth of complexity under control. Good tests are what keep a competent developer sane. You cannot build quality software without starting from tests. So if tests are boring you, the problem is your approach to engineering. Mature developers dont get bored chasing 100% coverage – they focus on meaningful tests that actually describe how the program is supposed to work.

gloosxon Nano Banana Pro

I dont think so. You cant train away a compression artifact that comes from the model's core architecture, LoRAs can smooth or hide artifacts, but some detail will be inevitably lost. You can try to hide artifacts but not remove them without retraining the whole model on RAW sensor data.

gloosxon Nano Banana Pro

At close-up inspection – 8x8 JPEG compression blocks are not going anywhere even with those "4k PNG images"

Seems like a fundamental flaw with image-models is that they will always output something resembling a JPEG

gloosxon Nano Banana Pro

Generated images still contain JPEG artifacts all over them.

We are not doomed yet - can pretty much reliably spot RAW image vs AI-generated image by just zooming in

gloosxon Google Antigravity

reads title

Wow was google researching some kind of anti-gravity device behind the curtains for real and then dropped it out of nowhere?

Ah damn, yet another ai-assisted-something. Crap.

gloosxon Show HN: Continuous Claude – run Claude Code in a loop

Ah yes, classic "increase test coverage for the sake of increasing test coverage".

Aligns with vibe-coding values well: number go up – exec happy.

gloosxon Windows 11 adds AI agent that runs in background with access to personal folders

CoPilot for BIOS

Integrated CoPilot chip, mandatory to install Windows

CoPilot for mouse movements. Just ask where you want to move your mouse next, agent does the job.

CoPilot which will entertain you while windows updates are installing

CoPilot-assistant to install Chrome browser

CoPilot for windows registry

Master-CoPilot to control all other copilots

CoPilot which will play games and watch movies instead of you, then give you 5-minute summaries to save your time

gloosxon AI World Clocks

anyone tried opening this from mobile? not a single clock renders correctly, almost looks like a joke on LLMs

gloosxon Yt-dlp: External JavaScript runtime now required for full YouTube support

You can get ad-free, fullscreen video on mobile right away by opening it from a chat message preview. You also just get the full-screen video, no way to minimise it and see Youtube UI, no ads visible. I was just wondering how it works and can you simulate that on the desktop.

gloosxon Yt-dlp: External JavaScript runtime now required for full YouTube support

The real problem is that YouTube built a model where the platform, not the creators, controls the money flow. They could have charged creators directly for hosting and left monetisation up to them, but by inserting themselves as the middleman, they gained leverage and authority over content itself. The "cost of hosting" is just the technical excuse for such centralisation.

gloosxon Yt-dlp: External JavaScript runtime now required for full YouTube support

I wonder how the whole thing works when I open a youtube video from a preview inside a chat application on my mobile phone.

It looks like the video loads and starts playing in some kind of in-app browser, but there is just full-screen video and nothing else. I also never faced any ads in this "mode" of playing a video, yet recently some strange things started happening where the playback would start together with an audio-track from the advertisement. The video itself would start playing but the sound would be replaced with the sound from ad which seemed very odd and much like a bug, only when advertisement audio track ends it will be replaced with audio track from the video itself.

I'm genuinely curious how is the whole playback process different when I watch a video from the Telegram preview, can I somehow achieve the same "just fullscreen video" kind of playback on the desktop as well? Does anyone have any insight?