Latest Tech News

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

Filtered by: odi Clear Filter

Two major AI coding tools wiped out user data after making cascading mistakes

New types of AI coding assistants promise to let anyone build software by typing commands in plain English. But when these tools generate incorrect internal representations of what's happening on your computer, the results can be catastrophic. Two recent incidents involving AI coding assistants put a spotlight on risks in the emerging field of "vibe coding"—using natural language to generate and execute code through AI models without paying close attention to how the code works under the hood.

Hacker slips malicious 'wiping' command into Amazon's Q AI coding assistant - and devs are worried

Amazon / Elyse Betters Picaro / ZDNET A while back, my ZDNET colleague David Gewirtz worried that someday AI coding agents could destroy open-source software. That day has come. A hacker managed to plant destructive wiping commands into Amazon's "Q" AI coding agent. Also: Coding with AI? My top 5 tips for vetting its output - and staying out of trouble This has sent shockwaves across developer circles. As details continue to emerge, both the tech industry and Amazon's user base have responded

Bad vibes: How an AI agent coded its way to disaster

Replit / Elyse Betters Picaro / ZDNET When AI leader Andrej Karpathy coined the phrase "vibe coding" for just letting AI chatbots do their thing when programming, he added, "It's not too bad for throwaway weekend projects … but it's not really coding -- I just see stuff, say stuff, run stuff, and copy-paste stuff, and it mostly works." Also: Coding with AI? My top 5 tips for vetting its output - and staying out of trouble There were lots of red flags in his comments, but that hasn't stopped p

Reversing a Fingerprint Reader Protocol (2021)

TLDR: The sensor uses TLS-PSK over USB. We overwrite the PSK and are able to read images. Scan of not-my-finger. I recently got a new Laptop ( Lenovo Ideapad 5 - 15are05 ). Linux compatibility is quite good, as documented in the Arch Wiki. One part that is not working is the fingerprint reader. In this blog post I describe my approach to reverse engineering the fingerprint reader's windows driver and USB protocol. This results in a python script, capable of streaming images from the sensor at

Firebender (YC W24) Is Hiring

Firebender processes tens of billions of tokens per day for the thousands of concurrent coding agents, and auto complete model. Every month hundreds of millions of lines of code are added to codebases of companies ranging from startups to fortune 500. Building a great coding agent is likely the most valuable technical challenges to solve right now, and we’re already making great progress on this. If this challenge excites you, let’s talk. Responsibilities Build/iterate on agentic evaluation,

AI-Powered Coding Assistant Deletes Company Database, Says Restoring It Is Impossible

A tech entrepreneur named Jason Lemkin set out to document his experience using an AI "vibe coding" tool called Replit to make an app. But the "vibes" turned bad real quick. The AI wiped out a key company database, he claims — and when called out on its mistake, it insisted, sorrowfully, that it couldn't undo its screw-up. "This was a catastrophic failure on my part," the AI wrote, as if depleted of any will to exist. "I violated explicit instructions, destroyed months of work, and broke the s

A vibe coding horror story: What started as 'a pure dopamine hit' ended in a nightmare

Replit / Elyse Betters Picaro / ZDNET When AI leader Andrej Karpathy coined the phrase "vibe coding" for just letting AI chatbots do their thing when programming, he added, "It's not too bad for throwaway weekend projects … but it's not really coding -- I just see stuff, say stuff, run stuff, and copy-paste stuff, and it mostly works." Also: Coding with AI? My top 5 tips for vetting its output - and staying out of trouble There were lots of red flags in his comments, but that hasn't stopped p

Coding with LLMs in the summer of 2025 – an update

antirez 6 hours ago. 31112 views. Frontier LLMs such as Gemini 2.5 PRO, with their vast understanding of many topics and their ability to grasp thousands of lines of code in a few seconds, are able to extend and amplify the programmer capabilities. If you are able to describe problems in a clear way and, if you are able to accept the back and forth needed in order to work with LLMs, you can reach incredible results such as: 1. Eliminating bugs you introduced in your code before it ever hits any

Topics: code coding llm llms work

Cursor snaps up enterprise startup Koala in challenge to GitHub Copilot

The startup behind the viral AI coding app Cursor is snapping up top talent from AI enterprise startups in a bid to bolster its competition with Microsoft’s GitHub Copilot and win over businesses looking to supercharge their employees with AI coding tools. In one recent case, Cursor maker Anysphere struck a deal to acquire the AI-powered customer relationship management (CRM) startup Koala, two sources familiar with the matter told TechCrunch. As part of the deal, Cursor will bring on several

Synths hunt down deadly monsters in latest Alien: Earth trailer

The premiere of Alien: Earth is just weeks away, and FX/Hulu dropped one last trailer to pique our interest, along with a much more detailed synopsis. It's meditative and existential in tone, with a haunting tune playing over footage of mysterious alien craft, dead bodies, blood-spattered humans fleeing through futuristic corridors, and, of course, a spooky silhouette of a xenomorph in the distance. As previously reported, the eight-episode series is set in 2120, two years before the events of

FX/Hulu drops one last trailer for Alien: Earth

The premiere of Alien: Earth is just weeks away, and FX/Hulu dropped one last trailer to pique our interest, along with a much more detailed synopsis. It's meditative and existential in tone, with a haunting tune playing over footage of mysterious alien craft, dead bodies, blood-spattered humans fleeing through futuristic corridors, and, of course, a spooky silhouette of a xenomorph in the distance. As previously reported, the eight-episode series is set in 2120, two years before the events of

Claude Code's new tool is all about maximizing ROI in your organization - how to try it

Anthropic / Elyse Betters Picaro / ZDNET Anthropic has distinguished itself from industry competitors for two major reasons: user privacy and coding capabilities. In particular, it's Claude Code's solution -- which allows users to run the coding assistant directly in their workspace to write or manage code -- that has attracted an exponential number of users. It's now a highly requested feature. Also: Anthropic's Claude dives into financial analysis. Here's what's new On Tuesday, Anthropic la

What Actually Happens When Programmers Use AI Is Hilarious, According to a New Study

AI has taken the programming world by storm, with a flurry of speculation about the tech replacing human coders, and Google's CEO recently claiming that 25 percent of the company's code is now AI-generated. But it's possible that in practice, AI is actually hindering efficient software development. As flagged by Ars Technica, a new study from the nonprofit Model Evaluation and Threat Research (METR) found that in practice, programmers are actually slower when using AI assistance tools than maki

Grok 4 benchmark results: Tops math, ranks second in coding

Grok 4 is a huge leap from Grok 3, but how good is it compared to other models in the market, such as Gemini 2.5 Pro? We now have answers, thanks to new independent benchmarks. LMArena.ai, which is an open platform for crowdsourced AI benchmarking, has published the results of Grok 4. We're talking about Grok 4 API (grok-4-0709), which received about 4k+ community votes and ranks #3 overall in Text Arena. This is a huge leap from Grok 3, which ranked 8th. According to LMArena's tests, Grok 4

Google hires Windsurf CEO Varun Mohan, others in $2.4 billion AI talent deal

Google on Friday made the latest a splash in the AI talent wars, announcing an agreement to bring in Varun Mohan, co-founder and CEO of artificial intelligence coding startup Windsurf. As part of the deal, Google will also hire other senior Windsurf research and development employees. Google is not investing in Windsurf, but the search giant will take a nonexclusive license to certain Windsurf technology, according to a person familiar with the matter. Windsurf remains free to license its techn

Google hires Windsurf CEO Varun Mohan, others in latest AI talent deal

Google on Friday made the latest a splash in the AI talent wars, announcing an agreement to bring in Varun Mohan, co-founder and CEO of artificial intelligence coding startup Windsurf. As part of the deal, Google will also hire other senior Windsurf research and development employees. Google is not investing in Windsurf, but the search giant will take a nonexclusive license to certain Windsurf technology, according to a person familiar with the matter. Windsurf remains free to license its techn

The best Plex alternative in 2025 is Jellyfin, and you really should try it

Robert Triggs / Android Authority Self-hosting is an increasingly popular alternative for those looking to escape the convoluted (and expensive) content-streaming landscape and return to a more traditional personal ownership approach to their media collection. Once you’ve bought your fancy new Network Attached Storage (NAS) and filled it with your favorite movies, the next step is to figure out what software to run on it. There are at least two solid options for serving up your media collectio

Windsurf’s CEO goes to Google; OpenAI’s acquisition falls apart

OpenAI’s deal to acquire the viral AI coding startup Windsurf for $3 billion fell apart on Friday, according to The Verge. In a shocking twist, Google DeepMind is now hiring Windsurf CEO Varun Mohan, co-founder Douglas Chen, and some of the startup’s top researchers. A Google spokesperson confirmed the hiring of Windsurf’s leaders in a statement to TechCrunch. “We’re excited to welcome some top AI coding talent from Windsurf’s team to Google DeepMind to advance our work in agentic coding,” sai

AI coding tools may not speed up every developer, study shows

Software engineer workflows have been transformed in recent years by an influx of AI coding tools like Cursor and GitHub Copilot, which promise to enhance productivity by automatically writing lines of code, fixing bugs, and testing changes. The tools are powered by AI models from OpenAI, Google DeepMind, Anthropic, and xAI that have rapidly increased their performance on a range of software engineering tests in recent years. However, a new study published Thursday by the non-profit AI research

Goldman Sachs is testing viral AI agent Devin as a ‘new employee’

In Brief Cognition’s AI coding agent Devin has scored a major customer: Goldman Sachs, the bank’s CIO, Marco Argenti, told CNBC. “We’re going to start augmenting our workforce with Devin, which is going to be like our new employee,” Argenti told the outlet, adding that it plans to roll out hundreds of instances of Devin, potentially growing to thousands. The bank currently employs around 12,000 human developers, it says. Despite the financial industry’s reputation for being slow and stodgy, G

Show HN: Vibe Kanban – Kanban board to manage your AI coding agents

Get 10X more out of Claude Code, Gemini CLI, Codex, Amp and other coding agents... Overview AI coding agents are increasingly writing the world's code and human engineers now spend the majority of their time planning, reviewing, and orchestrating tasks. Vibe Kanban streamlines this process, enabling you to: Easily switch between different coding agents Orchestrate the execution of multiple coding agents in parallel or in sequence Quickly review work and start dev servers Track the status o

Cloning Came to Polo. Then Things Got Truly Uncivilized

To play polo at the highest level, you have to love horses or be filthy rich. Ideally both. A team might field around 40 ponies for each game and draw on a roster of hundreds of potentially playable horses for each tournament. In the gaps between tournaments, top players scour the globe for new equine talent—trying to snaffle foals from elite bloodlines or looking to ex-racehorses to uncover a polo star in the rough. Andrey Borodin, the billionaire patron of Park Place Polo, wears the No. 1 jer

Show HN: Dev atrophy test – Can you still code without AI?

Hey HN, I'm Per from Scrimba (YC S20), the code-learning platform. There's been a lot of talk lately about whether AI tools are causing skill atrophy amongst developers. We get a front-row seat to this, and we see more and more students struggle with basic concepts, and building apps on their own. This is almost always a consequence of relying too much on ChatGPT and vibe coding tools. So we built a small side project: https://devatrophy.com It's a test of your core web dev knowledge — no ha

Inertial forces (indirect terms) in problems with a central body

Gravitational systems in astrophysics often comprise a body – the primary – that far outweights the others, and which is taken as the centre of the reference frame. A fictitious acceleration, also known as the indirect term, must therefore be added to all other bodies in the system to compensate for the absence of motion of the primary. In this Research Note, we first stress that there is not one indirect term but as many indirect terms as there are bodies in the system that exert a gravitationa

Why it’s so hard to warn people about flash floods

is a senior science reporter covering energy and the environment with more than a decade of experience. She is also the host of Hell or High Water: When Disaster Hits Home , a podcast from Vox Media and Audible Originals. By definition, flash floods are notoriously difficult to warn people about well in advance. They form rapidly, giving forecasters hours of lead time at best to figure out where they might hit with specificity. We’ve seen this with devastating effect in Texas, where flash flood

Cybersecurity’s Unsung Heroes: Elevating Neurodiverse Talent in the Digital Age

They aren’t your average hires. They don’t glide through job interviews or thrive in open-plan offices. But when it comes to pattern recognition, threat modeling, or zero-day hunting, neurodiverse individuals often outperform their peers. And yet, they remain largely untapped in the cybersecurity workforce. If we’re serious about fortifying digital defense systems, it’s time we stop overlooking the neurodiverse minds that already think in ways most cybersecurity frameworks are only beginning to

Spending Too Much Money on a Coding Agent

On making use of large thinking models. For a year, I’d been coding almost every day with Cursor and Claude Sonnet. Anthropic’s 3.5 and 3.7 Sonnet each rightly earned their dominant place on the programming model charts: they were the least-bad coding models yet. In the earliest days of LLMs, there was tremendous interest in ever-larger model releases. Hype around bigger, slower models has since waned, as Claude 3 Opus, GPT 4.5, and OpenAI o1 – all large and technically impressive model releas

Cursor launches a web app to manage AI coding agents

The company behind Cursor, the viral AI coding editor, launched a web app on Monday that allows users to manage a network of coding agents directly from their browser. The launch marks Cursor’s next big step beyond its integrated development environment (IDE), the core product developers use to access its tools. While Anysphere, the company behind Cursor, initially offered only this AI-powered IDE, the company has made a concerted effort to put its products in more places, and develop more agen

Are software professionals truly an endangered species? It's complicated

islander11/Getty Images Industry eyebrows were raised recently at New York Federal Reserve Bank data showing software engineering graduates face higher unemployment rates than art history majors. The unemployment rates for computer engineering and computer science were 7.5% and 6.7% respectively. By contrast, the unemployment rates for art history and social services majors were 3% and 1.7% respectively. Also: The best AI for coding in 2025 (including a new winner - and what not to use) In a