Latest Tech News

Stay updated with the latest in technology, AI, cybersecurity, and more

Filtered by: code Clear Filter

Read This Before You Trust Any AI-Written Code

We are in the era of vibe coding, allowing artificial intelligence models to generate code based on a developer’s prompt. Unfortunately, under the hood, the vibes are bad. According to a recent report published by data security firm Veracode, about half of all AI-generated code contains security flaws. Veracode tasked over 100 different large language models with completing 80 separate coding tasks, from using different coding languages to building different types of applications. Per the repor

Launch HN: Gecko Security (YC F24) – AI That Finds Vulnerabilities in Code

Hey HN, I'm JJ, Co-Founder of Gecko Security ( https://www.gecko.security ). We're building a new kind of static analysis tool that uses LLMs to find complex business logic and multi-step vulnerabilities that current scanners miss. We’ve used it to find 30+ CVEs in projects like Ollama, Gradio, and Ragflow ( https://www.gecko.security/research ). You can try it yourself on any OSS repo at ( https://app.gecko.security ). Anyone who’s used SAST (Static Application Security Testing) tools knows th

Carbon Language: An experimental successor to C++

Carbon Language: An experimental successor to C++ Why? | Goals | Status | Getting started | Join us See our announcement video from CppNorth. Note that Carbon is not ready for use. Fast and works with C++ Performance matching C++ using LLVM, with low-level access to bits and addresses Interoperate with your existing C++ code, from inheritance to templates Fast and scalable builds that work with your existing C++ build systems Modern and evolving Solid language foundations that are easy

Why AI researchers are getting paid like NBA All-Stars

Hello, and welcome to Decoder! This is Alex Heath, your Thursday episode guest host and deputy editor at The Verge. Today, I’m joined by Hayden Field, The Verge’s senior AI reporter. We’re talking about the AI talent wars and why some researchers are suddenly getting traded like they’re NBA superstars. If you’ve been paying attention at all to the AI industry this past month, you’ve likely seen stories about the hiring frenzy that’s been happening at Big Tech, research labs, and AI startups. Le

Following Up on the Python JIT

Following up on the Python JIT Please consider subscribing to LWN Subscriptions are the lifeblood of LWN.net. If you appreciate this content and would like to see more of it, your subscription will help to ensure that LWN continues to thrive. Please visit this page to join up and keep LWN on the net. Performance of Python programs has been a major focus of development for the language over the last five years or so; the Faster CPython project has been a big part of that effort. One of its subp

Proton releases a new app for two-factor authentication

Privacy-focused productivity tool company Proton released a new authenticator app today, allowing users to log in to services using dynamically generated two-factor authentication codes. The free app is available on all platforms starting today, including iOS, Android, Windows, macOS, and Linux. The app allows users to sync codes and accounts across devices. The company said that just like its other products, Proton Authenticator is open source and uses end-to-end encryption to protect user dat

I know when you're vibe coding

3 minute read I shouldn’t have to care about this. I don’t want to care about how someone’s code gets into the IDE. Whether you wrote it by hand, copied it from a forum, prompted an LLM, or ran a simulation where monkeys are given infinite time to produce the solution. I care about what gets merged into the codebase. When I click that “Approve” button, I’ve got only a few worries on my mind. Does it produce the correct outcome? Will people understand this next quarter? Will they be able to ch

Topics: care code don want write

Tracking source locations in the Futhark compiler

Posted on July 29, 2025 Futhark is a programming language meant for writing fast programs, but as is the case for every programming language meant for writing fast programs, it inevitably happens that a programmer will use it to write a program that is not fast. When this happens, the programmer will likely want to know why their program is not fast, and how to make it faster. A useful tool for answering such questions is a profiler - a tool that tells you how long the different parts of your p

Vibe code is legacy code

Despite widespread confusion, Andrej Karpathy coined "vibe coding" as a kind of AI-assisted coding where you "forget that the code even exists." Legacy code We already have a phrase for code that nobody understands: legacy code. Legacy code is universally despised, and for good reason. But why? You have the code, right? Can't you figure it out from there? Wrong. Code that nobody understands is tech debt. It takes a lot of time to understand unfamiliar code enough to debug it, let alone intro

Emacs: The macOS Bug

Emacs: The MacOS Bug The Context I have been recently roaming. Doing some Zig, doing some Go, some Janet. Some C integration. Should have focused on my project but life threw more at me than I could handle, so I sought… happy distractions. My experience with those technologies taught me new tricks and one day, when I needed some more distraction, I decided to debug something that had made me furious for years: Emacs jank. Whatever build I tried, whatever configuration I used, Emacs always r

Fast

fast Rarely in software does anyone ask for “fast.” We ask for features, we ask for volume discounts, we ask for the next data integration. We never think to ask for fast. But software that's fast changes behavior. Developers ship more often when code deploys in seconds (or milliseconds) instead of minutes. AI code complete means we can prototype in languages we're less familiar with. Real-time streaming makes remote work possible. Conversely, slow software limits us more tha

Flaw in Gemini CLI coding tool could allow hackers to run nasty commands

Researchers needed less than 48 hours with Google’s new Gemini CLI coding agent to devise an exploit that made a default configuration of the tool surreptitiously exfiltrate sensitive data to an attacker-controlled server. Gemini CLI is a free, open-source AI tool that works in the terminal environment to help developers write code. It plugs into Gemini 2.5 Pro, Google’s most advanced model for coding and simulated reasoning. Gemini CLI is similar to Gemini Code Assist except that it creates or

Google confirms it will sign the EU AI Code of Practice

Big Tech is increasingly addicted to AI, but many companies are allergic to regulation, bucking suggestions that they adhere to copyright law and provide data on training. In a rare move, Google has confirmed it will sign the European Union's AI Code of Practice, a framework it initially opposed for being too harsh. However, Google isn't totally on board with Europe's efforts to rein in the AI explosion. The company's head of global affairs, Kent Walker, noted that the code could stifle innovati

PlayerZero raises $15M to prevent AI agents from shipping buggy code

As Silicon Valley races toward a future where AI agents do most of the software programming, a new problem is created: finding the AI-generated bugs before they are put into production. Even OpenAI is dealing with such issues, a former employee has described. Newly funded startup PlayerZero has created a solution: use AI agents trained to find and fix problems before the code is put into production, the startup’s CEO and sole founder, Animesh Koratana, tells TechCrunch. Koratana created Player

Flickering lights could help fight misinformation

A group of Cornell computer scientists has unveiled what they believe could be a new tool in the fight against AI‑generated video, deepfakes and doctored clips. The watermarking technique, called “noise‑coded illumination,” hides verification data in light itself to help investigators spot doctored videos. The approach, devised by Peter Michael, Zekun Hao, Serge Belongie and assistant professor Abe Davis, was published in the June 27 issue of ACM Transactions on Graphics and will be presented b

Google will sign EU's AI Code of Practice

Google says it will sign the European Union’s new AI Code of Practice, which provides a framework for compliance with the EU’s AI Act. The act itself was passed in 2024, but its many provisions will take months to years to come into effect. The non-binding Code of Practice is a voluntary measure intended to help ensure that companies generally meet the obligations laid out by the Act in the meantime. In a blog post announcing Google’s participation, the tech giant shared some skepticism about t

Topics: act ai code eu europe

Most developers use AI in their daily workflows - but they don't trust it, study finds

fotograzia/Getty Images Programmers are using AI more than ever, but they don't like or trust the tools very much, according to the 2025 Stack Overflow Developer Survey. The survey of almost 50,000 developers found that 84% now use or plan to use AI tools in their workflow, up from last year (76%). Over half of professional developers (51%) use these tools daily. Also: The best AI for coding in 2025 (and what not to use) Such figures might suggest that programmers must love AI. However, only

Google says it will sign EU’s AI code of practice

Google has confirmed it will sign the European Union’s general purpose AI code of practice, a voluntary framework that aims to help AI developers implement processes and systems to comply with the bloc’s AI Act. Notably, Meta earlier this month said it would not sign the code, calling the EU’s implementation of its AI legislation “overreach,” and stating that Europe was “heading down the wrong path on AI.” Google’s commitment comes days before rules for providers of “general-purpose AI models

Topics: act ai code eu risk

For programmers, even as AI adoption climbs, trust wanes

fotograzia/Getty Images Programmers are using AI more than ever, but they don't like or trust the tools very much, according to the 2025 Stack Overflow Developer Survey. The survey of almost 50,000 developers found that 84% now use or plan to use AI tools in their workflow, up from last year (76%). Over half of professional developers (51%) use these tools daily. Also: The best AI for coding in 2025 (and what not to use) Such figures might suggest that programmers must love AI. However, only

CodeCrafters (YC S22) is hiring first Marketing Person

Connect directly with founders of the best YC-funded startups. In this role, you’ll own end-to-end marketing at CodeCrafters. CodeCrafters (YC S22) makes programming challenges for experienced software engineers. Besides YC, we're backed by the founders of Instagram and Dropbox. Our customers are experienced engineers at companies like Google, GitHub, Stripe, etc.

Playing with Open Source LLMs

Every 6 months or so, I decide to leave my cave and check out what the cool kids are doing with AI. Apparently the latest trend is to use fancy command line tools to write code using LLMs. This is a very nice change, since it suddenly makes AI compatible with my allergy to getting out of the terminal. Me, browsing HN from my cave (by Stable Diffusion) The most popular of these tools seems to be Claude Code. It promises to be able to build in total autonomy, being able to use search code, write

Why CI/CD Still Doesn't Include Continuous Documentation?

In my 15+ years as a developer, one of the most persistent headaches I’ve seen across teams is outdated documentation. I’ll admit it, I’ve shipped features and moved on without updating the docs. A month later, a new teammate is onboarding or someone is debugging an issue, and they run into a README or guide that no longer reflects reality. It’s frustrating for them and embarrassing for us. I’m certainly not alone in this habit. Maintaining documentation is often the last thing on a developer’

Structuring large Clojure codebases with Biff

Jacob O'Bryant | 28 Jan 2025 I've been making some progress on rewriting Yakread (a fancy reading app) from ~scratch and open-sourcing it in the process. Along the way I'm experimenting with potential new features for Biff, my Clojure web framework, which Yakread is built with. In particular I'm working on approaches for keeping Biff apps more manageable as the codebase grows: the original Yakread codebase was about 10k lines and was already getting pretty crufty. I've also learned some things

Stack Overflow data reveals the hidden productivity tax of ‘almost right’ AI code

Want smarter insights in your inbox? Sign up for our weekly newsletters to get only what matters to enterprise AI, data, and security leaders. Subscribe Now More developers than ever before are using AI tools to both assist and generate code. While enterprise AI adoption accelerates, new data from Stack Overflow’s 2025 Developer Survey exposes a critical blind spot: the mounting technical debt created by AI tools that generate “almost right” solutions, potentially undermining the productivity

Do variable names matter for AI code completion? (2025)

Do Variable Names Matter for AI Code Completion? When GitHub Copilot suggests your next line of code, does it matter whether your variables are named current_temperature or just x ? I ran an experiment to find out, testing 8 different AI models on 500 Python code samples across 7 naming styles. The results suggest that descriptive variable names do help AI code completion. The Experiment Each code sample was transformed into different naming conventions: Descriptive ( process_user_input )

Show HN: 433 – How to make a font that says nothing

433 is a font that masks visible text and replaces it with dots. I'm using it in the new version of Ensō along with Coffeeshop Mode. You can see it in action here: Beautiful. Why In short: because it's the simplest way to add Coffeeshop Mode to the app. (Also, because I was curious) (OK, mainly because I was curious, Dog mode) Longer version: The previous note on Ensō ended up on the front page of HN and gave us almost 400 TestFlight testers. This is way more than I had expected. In fact

LG Promo Codes: 20% Off | July 2025

LG makes some of the best TVs you can buy. Its OLED TVs in particular are perennial favorites at WIRED, with C-series models like the C4 (9/10, WIRED Recommends) providing among the best performance for your dollars on the market. LG is about way more than TVs of course. The Korean brand offers multiple products across the A/V landscape, from soundbars to Bluetooth speakers, along with a host of other products like home appliances, laptops, and more. Save 20% With Today’s LG Promo Codes If you

Show HN: 433 - How to Make a Font That Says Nothing

433 is a font that masks visible text and replaces it with dots. I'm using it in the new version of Ensō along with Coffeeshop Mode. You can see it in action here: Beautiful. Why In short: because it's the simplest way to add Coffeeshop Mode to the app. (Also, because I was curious) (OK, mainly because I was curious, Dog mode) Longer version: The previous note on Ensō ended up on the front page of HN and gave us almost 400 TestFlight testers. This is way more than I had expected. In fact

Anthropic is rate limiting Claude Code, blaming some users for never turning it off

Anthropic has introduced new weekly rate limits on its Claude Code tool for AI assistance with coding tasks. The move comes shortly after the AI company quietly began implementing rate limits on the Claude Code service, which is an agentic side of the AI chatbot that is capable of reading code, editing files, performing tests and pushing GitHub commits. According to a series of posts from Anthropic on X, these changes are in response to some users who have been running Claude Code "continuously

I designed my own fast game streaming video codec – PyroWave

Streaming gameplay from one machine to another over a network is a reasonably popular use case these days. These use cases demand very, very low latency. Every millisecond counts here. We need to: Send controller input from machine A to B over network B renders a frame on the GPU B encodes the frame into a bitstream B sends the result over a network to A A decodes the bitstream A displays image on screen Dopamine is released in target brain Every step in this chain adds latency and we wa

Topics: 32 bit codecs rate video