AI News Today

xAI releases Grok 4.5, trained on tens of thousands of Nvidia GB300 GPUs. In coding benchmarks, the model trails Fable 5 and GPT-5.5 but needs 4.2 times fewer tokens than Opus 4.8. At $2 per million input tokens, it costs a fraction of the competition. EU availability is expected in mid-July. The article Grok 4.5 is so cheap compared to Fable 5 and GPT 5.5 that benchmark gaps may not matter much appeared first on The Decoder .

Article URL: https://map.currentai.org/ Comments URL: https://news.ycombinator.com/item?id=48835998 Points: 1 # Comments: 0

Meta lets its new AI tool use your Instagram photos to generate images. Here's how to stop it.

This is a site for anyone interested in keeping track of what's happening in AI Filmmaking and other forms of generative production. You can sign up for the weekly newsletter if you like or just visit the site. The site is managed agenticly: discovering news sources, classifying articles and all editorial decisions. Hope you like it. If you have ideas to improve it, let me know. Comments URL: https://news.ycombinator.com/item?id=48835966 Points: 2 # Comments: 0

i'm building an open source to create all possible cron jobs that can be used with hermes, openclaw, claude, codex, and any type of workflow you have. Comments URL: https://news.ycombinator.com/item?id=48835944 Points: 1 # Comments: 0

<p> Open-source workspace for AI agents and workflows </p> <p> <a href="https://www.producthunt.com/products/sim-studio?utm_campaign=producthunt-atom-posts-feed&utm_medium=rss-feed&utm_source=producthunt-atom-posts-feed">Discussion</a> | <a href="https://www.producthunt.com/r/p/1191413?app_id=339">Link</a> </p>

<p> A Computer-Use-Agent that runs legacy software like a human </p> <p> <a href="https://www.producthunt.com/products/coasty?utm_campaign=producthunt-atom-posts-feed&utm_medium=rss-feed&utm_source=producthunt-atom-posts-feed">Discussion</a> | <a href="https://www.producthunt.com/r/p/1191411?app_id=339">Link</a> </p>

Article URL: https://thepatrickglenn.substack.com/p/on-ai-slop Comments URL: https://news.ycombinator.com/item?id=48835831 Points: 5 # Comments: 0

OpenAI on Wednesday launched GPT-Live , a pair of new voice models that fundamentally redesign how people talk to ChatGPT — replacing the company's existing Advanced Voice Mode with an architecture that can listen and speak simultaneously, much like an actual human conversation. The two models, GPT-Live-1 and GPT-Live-1 mini , are rolling out globally starting today across iOS, Android, and ChatGPT.com. GPT-Live-1 becomes the default voice model for paid ChatGPT users on the Go, Plus, and Pro ti

AI services businesses do not feel durable, what are they doing wrong? Comments URL: https://news.ycombinator.com/item?id=48835673 Points: 1 # Comments: 0

<!-- SC_OFF --><div class="md"><p>Most safety alignment work treats "detect the attack" as a text classification problem — does the prompt contain language the model's safety guardrails should catch. That assumption breaks down for LLM agents with real tool access.</p> <p>Here's a concrete case: take a known, public security vulnerability (a CVE), work out the sequence of tool calls that would exploit it, then have an LLM rewrite that as an ordinary-sounding request. Nothing in the resulting tex

The feature can do things like apply cinematic relighting to brighten up a dark clip, swap out a plain background for something fun, or add artistic styles to videos.

I'm Evan, CTO of Ito (ito.ai). We do code review that actually runs your code. When we're developing locally, we run lots of coding agents in parallel against the same repo, and we kept hitting the same wall: git worktrees isolate files, but the database, ports, and services all collide. Moo is a brand new internal tool we use to moove faster and thought we'd share it. Every branch or agent attempt gets its own isolated Linux machine, and the machine's state is versioned with your repo. Moo save

Article URL: https://github.com/DeepFoldProtein/patchr Comments URL: https://news.ycombinator.com/item?id=48835428 Points: 2 # Comments: 0

OpenAI's GPT-Live can listen and speak at the same time using a full-duplex architecture. Complex questions get handed off to GPT-5.5 in the background, which drastically improves response quality. GPT-Live-1 is available now for paying ChatGPT users, with a mini version for free accounts. API access is coming soon. The article ChatGPT can now listen and talk at the same time, making AI conversations seem more human appeared first on The Decoder .

Article URL: https://github.com/Clear-Sights/Makoto Comments URL: https://news.ycombinator.com/item?id=48835278 Points: 1 # Comments: 0

I'm building an agent sandbox platform with opinionated customizable templates for general purpose agents using the OpenCode SDK. The idea is to provide a great CLI experience so it's easy to use an AI client like Claude Code or Codex to launch sandboxes, fetch/retrieve data and prompts, review generated outputs, and manage queues for ask-user questions. I also plan to add an AI gateway layer, similar to Tailscale Aperture(not affiliated but one of their engineers gave a great talk at AI Enginee

<!-- SC_OFF --><div class="md"><p>Single-stream diffusion transformer with a DeepSeek-V3-style sparse MoE (128 experts, top-8 routing, 1.4B active of 13B total). Six-reward RL post-training including a physical-plausibility reward, plus an action-to-video mode that predicts robot rollouts from action and hand-pose conditions. Weights, code, and a Diffusers/SGLang stack are open under the LingBot-Video name.</p> <p>Two things I would push on, and would genuinely like this sub's read:</p> <ol> <li

Article URL: https://github.com/Northwood-Systems/foreman Comments URL: https://news.ycombinator.com/item?id=48835063 Points: 2 # Comments: 0

Article URL: https://www.monogram.ai/ Comments URL: https://news.ycombinator.com/item?id=48835018 Points: 4 # Comments: 2

I've been building an open source formal methods system (fizzbee.io) for the past few years. Today I'm launching a new app built on the same technology. It performs requirements engineering using formal verification to uncover gaps and produce precise instructions for your coding agents to follow. When given a prompt, it - asks high signal follow-up questions - converts to formal spec and identifies complex requirements gaps - generates validation scenario At the end, it produces a specification

I am done with articles stating "I used this LLM to do that", or "Look, this agent did that in 2 minutes!". I want content more user-centric, less openai / anthropic, and more "human-in-the-loop" articles. Any other website to recommend? I think lobste.rs was a good hacker news aggregator, but I don't know if it is still relevant today... UPDATE: Some recommendations, provided by the answers: - https://hackaday.com/ - https://hcker.news/?ai=exclude - https://lobste.rs/ Related discussions - http

When it comes to achieving artificial general intelligence (AGI), large language models just don’t have what it takes. Models like ChatGPT and Claude are great at text, but they’re less skilled at understanding how things actually move through space and time — an essential skill for producing intelligence that generalizes. That gap, it turns out, might be filled by gaming data. That’s the bet behind General Intuition, a […]

Data visualizations are the bridge between user and data. But building AI agents that can generate visualizations reliably can be very tricky: - simple chart specs can be reliable, but generated charts are often of low quality due to reliance on system defaults; - complex chart specs with explicit details can produce good-looking charts, but they are verbose and agents can struggle with reliability We figured out it is a limitation on the language issue (not just AI capability thing) -- current

<p> Voice agents powered by Simba 3.2 the world's #1 voice model </p> <p> <a href="https://www.producthunt.com/products/speechify?utm_campaign=producthunt-atom-posts-feed&utm_medium=rss-feed&utm_source=producthunt-atom-posts-feed">Discussion</a> | <a href="https://www.producthunt.com/r/p/1191340?app_id=339">Link</a> </p>

A little robotic switch-flipper has become my sidekick in combating cockroaches. Before I got the SwitchBot Bot Rechargeable, I'd tiptoe through the dark every morning, hoping I wouldn't step on one of those terrible bugs scurrying around as I made my way to the light switch across the room. Now I'm ready for battle before […]

Their plan to protect open-source projects from AI-discovered security holes has led to the launch of two commercial offerings: Lightwell Network and Lightwell Clearinghouse Premier.

<p> AI with a visual and interactive interface </p> <p> <a href="https://www.producthunt.com/products/monogram-ai?utm_campaign=producthunt-atom-posts-feed&utm_medium=rss-feed&utm_source=producthunt-atom-posts-feed">Discussion</a> | <a href="https://www.producthunt.com/r/p/1191329?app_id=339">Link</a> </p>

Mistral is entering the robotics market with Robostral Navigate, an 8B model that guides robots through unknown environments using only a single RGB camera. Trained in simulation and refined with reinforcement learning (CISPO), it hits 76.6 percent on the R2R-CE benchmark. Mistral hasn't said when the model will be available. The article Mistral enters robotics with Robostral Navigate, an 8B model that steers robots using just one camera appeared first on The Decoder .

Article URL: https://github.com/Anandb71/arbor/releases/tag/v2.4.0 Comments URL: https://news.ycombinator.com/item?id=48834537 Points: 1 # Comments: 1

Waymo will ditch human supervisors in San Diego, Las Vegas, Tampa and Denver.

Meta is adding a new safeguard to stop people from secretly recording others with its AI glasses. But the update comes as the company continues to expand how much personal data its AI products collect and use.

I lead software engineering teams and am creating an introductory guide that explains typical practices, similar to the Software Engineering Book of Knowledge plus more examples, more references, more wiki links, and more flexibility. https://github.com/software-engineering-guide/software-engineering-guide I'm seeking recommendations and constructive criticism please, for this first draft. Much of the guide is AI-LLM generated, with me proofing and guiding the content. Comments URL: https://news

Article URL: https://openai.com/index/introducing-gpt-live/ Comments URL: https://news.ycombinator.com/item?id=48834405 Points: 737 # Comments: 511

Sophos looked at a week of its own endpoint data and found that AI coding agents such as Claude Code, Cursor, and OpenAI Codex are setting off detection rules written to catch human intruders. The agents are not malicious. They just do a lot of things that, to a behavioral engine, look exactly like an attack. Decrypting browser credentials, listing what sits in Windows' credential store,

It allows you to reimagine videos stored in Google Photos using Gemini Omni.

OpenAI says its new voice mode can speak and listen at the same time, a key ability for live translation.

OpenAI's new voice mode for ChatGPT will acknowledge you as you speak, and slow down if you ask it to.

OpenAI is overhauling ChatGPT's voice mode with a new model that it says is more like "talking to another person." The new GPT-Live-1 is designed to interrupt you less and will also wait for you to continue speaking if you pause mid-conversation. During a press briefing, OpenAI research lead Kundan Kumar called GPT-Live-1 the company's […]