frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

My 3 SaaS are evaporating. RIP indie hacking (2018-2026)

https://twitter.com/gherget/status/2082784434726592780
2•gaborme•5m ago•0 comments

Excessive time spent online linked to stress and worse mood – study

https://www.theguardian.com/technology/2026/jul/29/excessive-time-spent-online-linked-to-stress-a...
2•giuliomagnifico•7m ago•0 comments

Cleaning Out Inboxes: TA488 Comes for Outlook with Another Half-Click Exploit

https://www.proofpoint.com/us/blog/threat-insight/cleaning-out-inboxes-ta488-comes-outlook-anothe...
1•romaniitedomum•7m ago•0 comments

Our Research Team Developed a More Efficient Audio Embedding Model

https://www.edgeimpulse.com/blog/how-our-research-team-developed-a-more-efficient-audio-embedding...
1•dansitu•7m ago•0 comments

Show HN: Claude Skill for ASD-STE100 – Simplified English

https://github.com/blagoySimandov/asd-ste100-writer-skill
2•blagoysimandoff•12m ago•0 comments

Copy Error in Apple's Developer Program Signup Screen

https://drive.google.com/file/d/1Pn5MNH-vaIKVOqZyueLQq_5a7psra9BL/view?usp=sharing
1•rcshubhadeep•12m ago•1 comments

Software Foundations Series

https://softwarefoundations.cis.upenn.edu/
1•iamanatom•13m ago•0 comments

Uefa to discuss boycotting World Cup in emergency meeting after FIFA bombshell

https://www.independent.co.uk/sport/football/fifa-world-cup-uefa-boycott-meeting-infantino-news-b...
1•vrganj•14m ago•0 comments

A fundamental flaw leaves LLMs strikingly vulnerable to attack

https://www.technologyreview.com/2026/07/30/1140927/a-fundamental-flaw-leaves-llms-vulnerable-to-...
2•joozio•15m ago•0 comments

OpenAI's Hacking Debacle Was a Human Mistake

https://www.wired.com/story/openais-hacking-debacle-was-a-human-mistake/
1•smurda•16m ago•0 comments

Need a Personal Injury Lawyer in Glasgow?

https://www.khanlaw.co.uk/
1•listingskdm•17m ago•1 comments

Show HN: A minimalist proxy for your local LLM cluster (~1100 lines)

https://github.com/robolamp/smol-llm-proxy
1•snowy_robolamp•19m ago•0 comments

The Chinese robot army transforming the UK's retail industry

https://www.bbc.co.uk/news/articles/c0jl8v23qwgo
1•edward•20m ago•0 comments

Parseword

https://www.parseword.com
1•bookofjoe•20m ago•0 comments

Typeless is stripping away my thought process

https://en.astrian.lol/think-dont-speak/
1•Astrian•21m ago•0 comments

Code Page 437

https://en.wikipedia.org/wiki/Code_page_437
1•tosh•21m ago•0 comments

When I failed the Microsoft interview

https://ochagavia.nl/blog/that-time-when-i-failed-the-microsoft-interview/
1•birdculture•23m ago•0 comments

"The Proximal Origin of SARS-CoV-2"

https://www.whitehouse.gov/lab-leak-true-origins-of-covid-19/
2•just_some_user•24m ago•1 comments

Mythos attack on 3rd-round PQC algorithm candidate puts it out of commission

https://arstechnica.com/security/2026/07/mythos-uncovers-crypto-weaknesses-that-went-unknown-for-...
3•throw0101a•26m ago•0 comments

k/simple

https://github.com/kparc/ksimple
1•tosh•29m ago•0 comments

BeHive – Open-source research engine that outputs structured claims, not essays

https://github.com/qa10devteam/behive
1•qadevteam•31m ago•0 comments

Origami research takes top prize at 2026 Regeneron ISEF competition

https://www.snexplores.org/article/2026-regeneron-isef-winners-origami
1•viscousviolin•32m ago•1 comments

Created Mono8 (PICO-8 style game engine built on MonoGame, .NET 8)

https://mono8games.itch.io/mono8
1•rurhdhdije•32m ago•1 comments

Coinbase moved to open models. Cut AI spend in half while increasing token usage

https://twitter.com/brian_armstrong/status/2070670644577280109
2•smurda•33m ago•0 comments

A Most Comfortable Dead End

https://monokai.com/articles/a-most-comfortable-dead-end/
1•monokai_nl•33m ago•0 comments

Frhog

https://www.frhog.com/?ribbit
1•krm01•34m ago•0 comments

In AI arms race, Swiss neutrality is double-edged sword

https://www.swissinfo.ch/eng/foreign-affairs/swiss-neutrality-is-double-edged-sword-in-ai-arms-ra...
1•TMWNN•36m ago•0 comments

Gpiozero Flow

https://bennuttall.com/blog/2026/07/gpiozero-flow/
21•benn_88•43m ago•3 comments

Atomarine: Nuclear Data Centers at Sea

https://atomarine.co/
25•ndr•44m ago•25 comments

GPT-5.6 SOL pushing 25h towards a goal

https://testflight.apple.com/join/vWrJPqk5
2•dreambiggg•47m ago•3 comments
Open in hackernews

GPT-5.6 SOL pushing 25h towards a goal

https://testflight.apple.com/join/vWrJPqk5
2•dreambiggg•47m ago

Comments

dreambiggg•47m ago
I build stuff in the in the JavaScript ecosystem. But now with this AI madness I said to myself: "How hard can it be to force GPT-5.6 Sol to do whatever I want but to keep it with guardrails so that it can do the stuff how I want it and not hallucinate on every turn?"

This is the process. This is a Swift app. I don't know shit about Swift. Never wrote a line of code. What I did was I had a chat using a skill(grill-with-docs). After almost 6 hours and a huge amount of questions, the feature that I wanted to be implemented was done. The result was two files with the common understanding after this long discussion and also almost 20 issues in GitHub.

Then I said to myself: "OK I have everything in place. Obviously I have more around to close the cycle and make sure codex has all the stuff he needs to do the job, verify the job, etc.

At this point all was in place: - ios/ipad connected to mac - /ask-matt and /implement skills present(search them online, they are incredible) - codex goal ready: "see Github issues and start the implementation. use /ask-matt when you don't know the implementation flow. use /implement to implement each issue."

And with that the process has started and after almost 25 hours the job was done and the result was, I must admit, impressive! Now as a developer I don't fucking know how to feel about this but the more I play with this type of stuff/tests, I get myself into a point where I'm thinking writing code by hand may not be my day to day job. Rather I will get to a point where all the knowledge that I have, I will use it to have a long discussion with the LLM. We get the two of us acknowledging and defining everything that needs to be defined but not on a technical level but rather on how it should look and work and behave.

After all of this, one session, many sessions, have all these decisions on markdown files and then create tasks/issues so that the agent can take one by one and implement them overnight. Then you look at the result. Fucking hell this freaks me out to be honest. Maybe because I'm a freak control and maybe because I need to really understand the code. But now with this AI madness and everyone's saying, "Yeah we can do this in one week," I feel like as a developer I'm forced to go into the same journey or be left out. I guess I have so much to say about this but I wanted this post where I fucking complain about my fears and stuff regarding this AI madness and where things are going. If you read this until the end, I hope you have an amazing day. If you feel like trying the app from Test Flight, go ahead. Your feedback will be appreciated. Try to be nice and kind if you want. If you don't want to then I don't know. Thank you nevertheless.

saidnooneever•29m ago
i know its not nice but i would recommend to view work as a means to pay bills and find joys in a more controllable setting, like hobby projects.

this will make it easier to accept change, like this AI/LLM shift, and accepting that it wont bother you. if it doesnt bother you negativity does not occupy or cloud your mind and you can explore new ways to enjoy the new workflows you are needing to do, whislt knowing your upper mgmt cant have another brainfart to take away the joys in your life by changing up your tools or work.

i know its not an answer you asked for, but i went through similar dread. my work is developing prototypes in code so this is an obvious job that will surely die.

dreambiggg•24m ago
Well, I find joy in everything I do related to coding, even though I don't write the code myself. What I'm trying to say is that I'm not against AI stuff, but maybe I'm against how it's presented on social media—where it seems like everyone can do everything, like, yesterday. Everything is so easy that you don't really need developers, or even worse, anyone can build anything in a week.

But other than all of this BS, I still enjoy doing this type of work. Now with AI, I feel like I'm doing all the ideas I never got to because I didn't have the time, or the skill wasn't there for a given language, tool, and so on. With AI, I find myself willing to try more stuff and/or do more stuff.