News for hackers. Stuff that matters.

pacman1337

Re:Genie 4 on Tuesday October 27, 2026 @12:43PM (#50261868)
this is the gta6 I am was hoping for
Re:Nvidia discusses $2k-per-GPU annual fee for software on Monday October 26, 2026 @04:55PM (#50250805)
because NVIDIA can force them
Re:Type Server Protocol on Monday October 26, 2026 @04:40PM (#50250602)
languages with dynamic types will be obsolete thanks to AI, you heard it here first. It is just introduces more errors, with AI you want formal checking, sure static types are not that, but it is a small piece of it. If 10 percent of errors are types, with a typed language keeping everything the same you have 10 percent less errors, simple as that.
more help find people so they can start pointing fingers at each other
Re:Chapter 1 of Ways of Seeing by John Berger on Monday October 12, 2026 @11:30AM (#50057870)
I agree. Just scanned it and not getting anything of this word salad. What is the main point, is it anything significant? Some of the pictures are interesting that is about it.
Re:Unikernels were hard. key word: were on Sunday October 11, 2026 @07:53PM (#50049439)
The operating system and user programs including the application is the application. Worrying if it is unikernel or kernel is irrelevant at this point.
Re:Kimi K3 is now live on Thursday July 16, 2026 @02:45PM (#48938534)
Likely won't improve much. They trained on every text already.
Re:Amazon aims to raise $25B from bond sale on Wednesday July 08, 2026 @04:56PM (#48837294)
what interest rate are they offering?
Re:How Kalshi Infects the News on Monday July 06, 2026 @01:47PM (#48808050)
Adults should have the freedom to gamble if they feel like gambling. I don't like the government trying to daddy me. If I want to blow my money on some stupid bet that is my problem. As long as the rules are clear and fair then it should absolutely be legal. CNN not disclosing is bad, but that has nothing to do with gambling in general.
Re:Launch HN: Manufact (YC S25) – MCP Cloud on Thursday July 02, 2026 @11:56AM (#48763433)
stopped using mcp and mostly using skills now. can't understand what this product does or how it could help me.
Re:A Herculaneum scroll has been read for the first time on Thursday June 25, 2026 @03:45PM (#48678297)
Where is the direct English translation? I don't care about anything else.
Re:.gitignore Isn't the Only Way to Ignore Files in Git on Friday June 19, 2026 @03:53PM (#48602538)
another way is to use shortcuts
Good point, I didn't consider that.
Companies should just stop providing AI to police states like Germany. Apparently their citizens believe anything they read.
First comment that makes sense. Of course AI needs to be verified. Of course it gives bs answers. The answer is to educate the public not try to daddy then. I hope AI is completely banned from Germany for their protection since apparently their government doesn't believe its people know how to question AI.
Re:Claude Fable 5 on Tuesday June 09, 2026 @02:33PM (#48465478)
Yeah I noticed that too. For 98% of tasks I get same results with DeepSeek, it is starting to just be a branding game. It is incredible how marketing can get someone to pay 100x for same thing you can get for 1x.

This is why Claude Code just doesn't make sense to me. I need an agent that can plan using Opus and execute using DeepSeek or something else.

This exactly the issue. Anyone with common sense knows that social media have many negative effects, attention span, negative beliefs and ideas, addiction, etc. Just like inhaling smoke into your lungs can't be healthy right? But then comes so called scientist who say well since we thought the the universe was Euclidean but now we see it is non-Euclidean what is common sense is not common sense, there is no common sense. We need some sort of impossible proof to what was originally common sense.
this is why science credibility is going down, what we call science is abused. This is like saying smoking has no evidence it causes harm.
propaganda
The past making quality clothing was difficult, cutting it right, sourcing the right patterns designs materials, stitching it took care etc. In our world making quality clothing should be easy with all the technology but what we see in bad quality that you wash a few times and it is trash. It is uglier designs than in the past etc. It makes no sense. It is like a conspiracy where people don't want to sell quality clothes at a fair price. Like all companies got together and decided we will sell crap clothes at cheap prices and good clothes at extortion prices. There is zero correlation with actual costs.
Re:The current AI pricing was always going to go away on Friday May 22, 2026 @12:00PM (#48237687)
I get similar results for deepseek and opus but opus is way faster. I guess deepseek streams thinking and makes it slower?
Re:OpenAI Is Preparing to File for an IPO Soon on Thursday May 21, 2026 @05:16PM (#48228992)
A lot of people don't understand that creating a startup is not about having the best product in the market. It is about convincing people to part with their money and give it to them to control and gain power with, through having people work on what you want them to work on. Only requirement is you create a story of why they should give you money. Even if it is the dumbest idea and you will be lose all it in 3 years. If you can tell a convincing story you will still make yourself and your friends super rich.
wasn't there something similar with people tracking celebrities planes?
I need skill to block ads
Re:One item purchased, ten emails on Wednesday April 08, 2026 @03:05PM (#47694755)
pretty soon we won't be reading books or emails. We have have AI filters. Maybe we won't even be reading our friends or partners message either, mostly fluff anyways?
The irony no one is talking about: AI makes quality code worse. Was bad enough already so imagine it now. I am expecting many more services to drop from 3 nines to 1 nine.
Re:Edge.js: Run Node apps inside a WebAssembly sandbox on Tuesday March 17, 2026 @08:17PM (#47420147)
Not sure I understand the use cases. I'm guessing people want to run unsafe nodejs code. Either cloud providers like for lambdas or on personal computer for AI coding? On cloud why is this better than firecracker, firecrack can run any programming language this is just nodejs so already useless for that use case?

For personal computer, people worried about the spin up time of docker? I think that is more of a tooling issue where you spin up one instance and run multiple jobs.

What am I missing? What are actually real use cases where this would be better?

Re:I have written gemma3 inference in pure C on Thursday January 29, 2026 @09:49AM (#46810917)
Anyone using this model for something useful? For now I only have use cases for top performing models...
Re:Owners, not renters: Mozilla's open source AI strategy on Tuesday January 13, 2026 @12:34PM (#46604459)
You probably like watching ads because Firefox is only browser you will have a true ad free experience. Unfortunately Firefox is slower than chrome has less support for audio, copy pasting is broken etc. So I use both depending what I am doing.
Re:Mistral launches OCR 3 – 74% win rate over OCR 2 on Friday December 19, 2025 @09:08PM (#46333100)
what prompt are you using?

Older »