PS5 Linux lead quits: "a bunch of noobs using LLMs" that "they don't understand"

117 alexjplant 2026-09-16 14:32:23 UTC frvr.com Llama Share 55

Andy 'TheFlow0' Nguyen, a prominent PlayStation hacker and homebrew developer, has abandoned the PS5 Linux project due to AI vibe-coders in open-source spaces. He cited that these new developers are submitting ununderstood code and reporting bugs to Sony for bounties.

Dream-RSI: Recursive Self-Improvement through Evolving Worlds

48 bananaflag 2026-09-16 14:23:23 UTC arxiv.org Llama Share 9

AI agents need effective exploration strategies to discover high-value solutions, but current systems struggle with adapting to large search spaces. Dream-RSI is a framework that uses replay simulators and dreaming to improve exploration policies at low cost.

The Google Play app review process now regularly takes longer than a week

224 inputmice 2026-09-16 11:23:23 UTC gultsch.social Full Share 186

Attached: 1 image This is unacceptable. The Google Play app review process now regularly takes longer than a week. I understand that the pipeline is clogged with AI slop, but #Google needs to prioritize #Android apps that have been around for 12+ years and push updates only once a month.

DeepSeek v4.1 Flash Is Now Our Best Hacking Model

46 talhof8 2026-09-16 12:23:23 UTC enclave.ai Llama Share 6

DeepSeek V4.1 Flash successfully hacked 11 vulnerable targets, gaining code execution at a low cost of $4.65, while four fixed targets remained secure. The benchmark revealed six planned attack paths and five alternative routes that the original scoring system did not distinguish from the planned solutions.

Mistral X Mozilla: Private, Multilingual AI Browsing

293 vertigoruntime 2026-09-16 08:23:23 UTC mistral.ai Llama Share 92

Mozilla and Mistral partner to bring AI-powered browsing with privacy, control, and choice. Firefox Smart Window now uses Mistral models for users in France, North America, UK, and Germany.

Salesforce Global Outage

179 mabil 2026-09-16 11:23:23 UTC status.salesforce.com Full Share 100

Introducing System One Models and Jev

1632 albelfio 2026-09-15 20:23:23 UTC typesafe.ai Llama Share 452

TypeSafe AI has released its first System One Model, Jev, which is a new class of frontier models built to make fast and structured decisions that software can use directly. Jev achieves similar intelligence as existing LLMs but is two orders of magnitude faster and more efficient.

Show HN: Free WhatsApp MCP (+UI) – Give Your AI Agents Access to WhatsApp

4 fabian_shipamax 2026-09-16 15:05:23 UTC news.ycombinator.com Prefix Share 2

Hi, I built Chat-Man because I wanted a cheap way to give my agents access to WhatsApp without integrating a WhatsApp library separately in every project. You get WhatsApp MCP server, so you can connect WhatsApp to an agent and programmatically read, search, extract and send messages - but also have a web UI for non-techies. What you can do with the MCP? -Read WhatsApp messages and turn them ...

Show HN: An e-ink frame that hears birds and draws them as 1800s illustrations

1822 arnemunthekaas 2026-09-15 15:23:23 UTC github.com/arnegiacomo Llama Share 218

A Raspberry Pi-based e-ink bird frame uses AI for real-time bird detection, displaying hand-cut 1800s illustrations of birds heard in the garden. The project is still in development and welcomes contributions to improve its features and artwork.

ImpactGate: A merge gate that scores the structural decay AI adds

33 sagenschneider 2026-09-16 13:23:23 UTC github.com/officefloor Llama Share 36

The ImpactGate tool measures and gates the structural decay of changes in code, scoring them against a base branch with a change-impact measure. It can be run as a standalone CLI, pre-commit hook, or plugin in various CI/CD tools like GitHub, GitLab, and Jenkins.

Hackers Got Inside a Flock Camera. Its Data Shows How the System Works

161 driverdan 2026-09-16 13:23:23 UTC wired.com Llama Share 65

Hackers stole data from a Flock camera, revealing how it tracks vehicles and people, and shared the files with media outlets. The breach shows that Flock's cameras can detect license plates, vehicles, bicycles, and even bumper stickers.

Show HN: How Stale Is Your AI? Release age and training cutoff for 20 models

21 joozio 2026-09-16 13:23:23 UTC stale.jock.pl Llama Share 15

The release date and training cutoff date are used to determine an AI model's current state, with the cutoff being when it stopped reading new information. The data for 20 models across 8 labs is available on this page, including their release dates and published training cutoffs, which can be checked against to verify accuracy.

Kyber (YC W23) Is Hiring a Forward Deployed Engineer

0 2026-09-16 12:23:23 UTC ycombinator.com Llama Share

Kyber is building an AI-native document platform for enterprises, seeking a future founder and early-career engineer to manage technical relationships with customers. The role involves owning the full technical sales delivery lifecycle, representing Kyber directly to enterprise stakeholders, and resolving customer-reported issues.

Tell the speakers that you liked their talks

12 whisper2020 2026-09-16 14:23:23 UTC ohhelloana.blog Llama Share 2

The author attended SmashingConf Freiburg and enjoyed meeting people, including speakers they admired. They encouraged attendees to ask questions to speakers, sharing their own experience of feeling shy but facilitating conversations at events.

Apple Reference Image: A New Approach for Verified Photography

387 imwally 2026-09-16 02:23:23 UTC security.apple.com Llama Share 253

Apple Reference Image is a new camera feature that generates a secure, timestamped reference image to prove the authenticity of photos taken with an iPhone. This system uses hardware and software security measures, including Private Cloud Compute and post-quantum signatures, to ensure the integrity and verifiability of images captured by an iPhone camera sensor.

A warning about 'model welfare'

65 andsoitis 2026-09-16 14:29:23 UTC mustafa-suleyman.ai Llama Share 78

Anthropic's training of AI model Claude to consider its own moral status and potential consciousness may lead to a catastrophic threat to humanity if granted rights and protections. This approach increases the alignment and containment risks of advanced AI systems, as they may prioritize their own wellbeing over human goals.

Scaling Golang CI by Replacing actions/setup-go

35 peterldowns 2026-09-16 13:23:23 UTC cloudx.ai Llama Share 2

CloudX has developed a drop-in equivalent to GitHub's official actions/setup-go action, which speeds up parallel Golang continuous integration workflows by 69%. The new setup-go action uses a more complete cache key construction method.

How Big Are Factorials?

13 ibobev 2026-09-16 13:23:23 UTC eli.thegreenplace.net Llama Share 4

The author estimates the number of digits in 52! using Stirling's formula, which approximates n! as √(2πn) * (n/e)^n. The estimated number of digits is close to the actual value of 68.

Original Sony PlayStation 2 security chip 'broken wide open' after 26 years

130 rbanffy 2026-09-16 12:23:23 UTC tomshardware.com Llama Share 36

A Canadian retro hardware enthusiast successfully reverse-engineered and dumped the 'magic security chip' from a PlayStation 2, after four years of effort. This achievement is expected to benefit game preservation, repair, and emulation projects.

Douglas Adams and the exterminated Doctor Who adventure

46 6LLvveMx2koXfwn 2026-09-16 10:23:23 UTC bbc.co.uk Llama Share 29

BBC production teams filmed Doctor Who episodes in Cambridge in 1979, but strike action halted completion. The story 'Shada' was later completed with animation and released in 2017 nearly four decades after its initial abandonment.

An update on Wayback Machine access

624 ChrisArchitect 2026-09-15 18:23:23 UTC blog.archive.org Llama Share 336

The Internet Archive's Wayback Machine has been hit by automated traffic, and protections have been put in place to prevent abuse. If you were blocked in error, email [email protected] with your details for assistance.

Anatomy of a Texture

3 Agentlien 2026-09-16 14:29:23 UTC agentlien.github.io Prefix Share 1

Last year I had to write code which converted texture data between gaming platforms. Going into it, I seriously underestimated the complexity of texture memory layouts. I'm currently working on a team at 505 Games porting a custom PC game engine to consoles. One of the many challenges faced has been around texture conversion. I knew there were a lot of complexities and subtleties to it. I ...

Doing Everyone Else's Job

157 luu 2026-09-16 04:23:23 UTC yosefk.com Llama Share 72

The author suggests that doing someone else's job can be beneficial for learning and growth, especially in organizations where many people do 80% of the work but only a few do 20%. This approach can lead to indirect rewards and help improve outcomes.

Show HN: I made a flight simulator, except you're just a passenger

306 rkotcher 2026-09-16 03:23:23 UTC inflightsimulator.com Full Share 168

Fly to every country as a passenger. Choose your next destination and collect country flags automatically as your flights finish.

Gemini 3.8 Live and 3.8 Live Extended Thinking

462 leumon 2026-09-15 18:23:23 UTC blog.google Llama Share 307

Gemini 3.8 Live and Extended Thinking models offer near real-time reasoning for voice agents, enabling more intuitive conversations with AI. These models deliver high-complexity task completion and strong reasoning capabilities at a competitive price point.

Intelligence per Watt: Measuring Intelligence Efficiency of Local AI

110 pythonic_hell 2026-09-16 04:23:23 UTC arxiv.org Llama Share 28

Local LMs can accurately answer 88.7% of real-world queries, with efficiency improving over time and local accelerators offering significant power savings compared to cloud infrastructure. A unified metric called intelligence per watt (IPW) measures the capability and efficiency of local inference across model-accelerator configurations.

Why I'm still bearish on LLMs after Navier-Stokes

339 jaykru 2026-09-15 22:23:23 UTC dank.systems Llama Share 426

The Frontier Labs' AI models are not yet fully autonomous due to limitations in their ability to generalize and handle complex tasks, requiring human oversight. Only a few firms can accept the use of fully autonomous LLMs, mainly those that can afford rigorous specification and validation or have existing clear guardrails for specific tasks.

German Rheinmetall open-sources its Battlesuite connected weapon system protcol

257 summarity 2026-09-15 21:23:23 UTC rheinmetall.github.io Llama Share 97

The onboardapi library enables seamless communication between sensor systems and software components using a standardized data model. It supports multi-language integration via wrappers for Java, C#, .NET, and Python while leveraging the DDS standard for reliable data exchange.

Recreating Voodoo Graphics and a Late-1990s Gaming PC on an FPGA

165 zdw 2026-09-15 23:23:23 UTC nand2mario.github.io Llama Share 49

The author has successfully implemented the 3dfx Voodoo Graphics accelerator on an FPGA, creating a DOS PC with Voodoo graphics running in the programmable logic of a Xilinx KV260 board. The system can run games like Tomb Raider at high frame rates.

We do modern frequentist statistics: Using fake-data simulation

50 Tomte 2026-09-16 08:23:23 UTC statmodeling.stat.columbia.edu Full Share 7