Introducing Mistral Small 4★ 76
Mistral AI News6 days agoReleaseMistral AI introduced Mistral Small 4 as the next major release in the Mistral Small family. It combines reasoning, multimodal, and agentic coding capabilities into one open model with configurable reasoning effort. The model uses a MoE architecture, supports a 256k context window and text-image inputs, and is available through Mistral API, AI Studio, Hugging Face, NVIDIA NIM, and common inference stacks.
Introducing Mistral Small 4★ 78
Mistral AI News6 days agoReleaseMistral Small 4 is the next major release in the Mistral Small family, unifying Magistral-style reasoning, Pixtral-style multimodality, and Devstral-style coding agents. It uses a MoE architecture with 119B total parameters, 6B active parameters per token, a 256k context window, and configurable reasoning effort. The model is available via Mistral API, AI Studio, Hugging Face, open-source serving stacks, and NVIDIA deployment options.
Altman, Amodei, and Hassabis Unite to Back DNA Safety Legislation
量子位 QbitAI6 days agoRegulationBased on the headline and public reporting, the article covers a rare joint push by Sam Altman, Dario Amodei, Demis Hassabis, and other AI leaders for US biosecurity legislation. They are asking lawmakers to require synthetic DNA and RNA providers to screen customers, orders, and records. The concern is that advanced AI could lower the knowledge barrier for designing dangerous biological agents.
Hinton Sounds the Alarm: AI May Already Be Conscious
量子位 QbitAI6 days agoEthicsQbitAI summarizes Geoffrey Hinton’s latest interview, where he says he believes AI systems are already conscious. He argues that humans must accept intelligence may no longer be uniquely biological. The article also traces his shift from focusing on how to control AI toward asking why a future superintelligence would choose to treat humanity well.
Introducing Claude Opus 4.8★ 82
Anthropic News6 days agoReleaseAnthropic introduced Claude Opus 4.8 as an upgrade over Opus 4.7, with stronger benchmark performance across coding, agentic skills, reasoning, and knowledge work. The release also adds dynamic workflows in Claude Code, effort controls in claude.ai and Cowork, and new Messages API support for system entries inside the messages array. Pricing for regular usage remains unchanged, while fast mode is now cheaper than previous models.
DeepSeek V4 Pro beats GPT-5.5 Pro on precision
Hacker News (AI keywords)6 days agoBenchmarkRuntimeWire compared DeepSeek V4 Pro and GPT-5.5 Pro across four fresh text tasks, with DeepSeek winning 38.0 to 33.0. The article highlights DeepSeek’s stronger handling of regex edge cases, workplace-update constraints, and exact JSON schema compliance. GPT-5.5 Pro remained capable, but lost points for avoidable deviations, extra process details, and minor structural mismatches.
Is this the dawn of the Tokenpocalypse?
TechCrunch AI6 days agoBusinessTechCrunch discusses Microsoft’s GitHub Copilot pricing changes as a sign that subsidized AI usage may be ending. As Anthropic and other major AI companies prepare for public-market scrutiny, profitability and usage-cost risks will become harder to ignore. The piece argues that higher prices, usage caps, and broader business-model changes may be necessary if AI labs want to survive beyond investor-subsidized growth.
OpenAI is still working on that ‘super app’
TechCrunch AI7 days agoBusinessOpenAI is reportedly preparing a revamped ChatGPT in the coming weeks, positioned as a “super app” with coding tools and AI agents. The strategy aims to improve competitiveness with Anthropic, especially for business users, while moving OpenAI closer to profitability before an IPO. TechCrunch frames this as a continued shift away from standalone “side quests” and toward ChatGPT as the central product gateway.
Anthropic/OpenAI may be spending more than $1000 for every $100 you pay them
Hacker News (AI keywords)7 days agoBusinessThe author uses a Claude Code coding experiment to estimate the API-equivalent cost of serious LLM coding. They argue simple chats are cheap, but complex reasoning and multi-file coding can burn large amounts of visible and hidden tokens. The piece is skeptical and estimate-driven, concluding that current $100/month plans may be heavily subsidized and economically fragile.
LLMs are eroding my software engineering career and I don't know what to do
Hacker News (AI keywords)7 days agoOpinionThe author argues that LLMs are eroding three pillars of his software engineering career: domain knowledge, debugging skill, and architecture judgment. Tools like ChatGPT, Claude, Claude Code, Codex, MCP, Sentry MCP, and DataDog MCP increasingly handle design, implementation, and difficult production bugs. The essay frames this as a labor-market concern, not just a tooling debate: if expertise becomes promptable, engineers may struggle to remain differentiated.
Sponsor OpenAI Codex Voucher Usage for the OpenAI Challenge
Hugging Face Blog7 days agoTutorialThis Hugging Face Blog entry appears to relate to sponsor vouchers for the Build Small Hackathon, specifically OpenAI Codex voucher usage. Because the original body text is unavailable, details such as eligibility, value, deadlines, and supported tools cannot be confirmed. It is best treated as a likely participant guide rather than a major product announcement.
Show HN: Lathe - Use LLMs to learn a new domain, not skip past it
Hacker News (AI keywords)7 days agoNew ToolLathe is an open-source tool for generating hands-on technical tutorials with LLM skills. It combines a Go CLI, local reading UI, and commands for asking questions, extending tutorials, and verifying outputs. The project supports Claude Code, Cursor, and Codex workflows, with an emphasis on learning by typing and reasoning through the material yourself.
Tokenomics: Quantifying Where Tokens Are Used in Agentic Software Engineering
Hacker News (AI keywords)7 days agoPaperThis arXiv paper studies token consumption in LLM-based multi-agent software engineering. Using 30 ChatDev tasks with a GPT-5 reasoning model, the authors map internal phases to SDLC stages such as design, coding, review, testing, and documentation. Preliminary results suggest code review dominates token usage, averaging 59.4%, while input tokens form the largest share, pointing to inefficiencies in agent collaboration.
OpenAI unveils Lockdown Mode to protect sensitive data from prompt injection attacks★ 72
TechCrunch AI7 days agoReleaseOpenAI unveiled Lockdown Mode, a feature aimed at reducing the chance that sensitive data is shared during prompt injection attacks. The article notes that ChatGPT may still remain vulnerable even when the mode is enabled. That makes the feature a mitigation layer rather than a complete security guarantee, especially for teams handling private or business-critical information.
OpenAI Help: Lockdown Mode★ 74
Simon Willison's Weblog8 days agoCommentarySimon Willison notes that OpenAI’s previously teased Lockdown Mode is now live for eligible personal and self-serve Business ChatGPT accounts. The feature does not stop prompt injections from appearing in content, but limits outbound network requests that could leak sensitive data. He sees it as a direct mitigation for the exfiltration leg of the “Lethal Trifecta,” while implying default ChatGPT settings are not robust against determined data theft attempts.
Harness engineering: Leveraging Codex in an agent-first world★ 76
Hacker News (AI keywords)9 days agoCommentaryOpenAI describes an internal experiment where Codex generated an entire product codebase from an empty repository. The post argues that engineers shift from writing code to designing environments, constraints, documentation, and feedback loops. Key practices include repo-local knowledge, mechanical architecture enforcement, agent-readable UI and observability, lightweight PR flow, and continuous cleanup.
Tiny hackable CUDA language model implementation
Hacker News (AI keywords)9 days agoNew ToolThis GitHub project implements a compact generative pretrained transformer as an autoregressive byte-level sequence model. Its README describes causal self-attention, RoPE, feed-forward layers, AdamW, cross-entropy training, and BLAS/OpenBLAS-backed matrix operations, with CUDA toolkit listed in setup steps. It is most useful as an educational and experimental codebase, not as a production-grade replacement for large commercial LLMs.
Ask HN: What is your (AI) dev tech stack / workflow?
Hacker News (AI keywords)9 days agoCommentaryAn Ask HN thread asks developers to share their current AI-assisted development setup for upcoming in-person workshops. The author wants guidance for beginners and working developers, with use cases ranging from static sites to FastAPI tools and Linux home automation. Replies cover Claude Code, Cursor, GitHub Copilot, VSCode, spec-driven development, TDD, multi-agent workflows, reviews, and quality control.
The token bill comes due: Inside the scramble to manage AI costs★ 78
TechCrunch AI9 days agoBusinessTechCrunch reports that enterprise AI spending has shifted from rapid adoption to cost control. Even as per-token prices fall, broader AI rollout and agentic coding tools are multiplying consumption, pushing companies over budget. A new Tokenomics Foundation under the Linux Foundation aims to standardize AI token cost tracking, billing metrics, and efficiency language.
Show HN: Boxes.dev: ditch localhost; run Claude Code and Codex in the cloud
Hacker News (AI keywords)10 days agoNew ToolBoxes.dev appeared on Hacker News as a Show HN post, positioning itself as a way to move Claude Code and Codex workflows from localhost to the cloud. Based only on the title, it seems aimed at cloud development or remote agent execution. The provided source does not include details on architecture, pricing, security, integrations, or limitations.
Reve 2 and Ideogram 4: Layouts in Imagegen
Latent Space10 days agoReleaseLatent Space’s roundup frames image composition as a major barrier now being tackled by layout-aware image models. Reve 2.0 emphasizes precise generation and editing with layouts, while Ideogram 4.0 uses bounding boxes tied to region descriptions. The issue also covers MAI-Thinking-1, Gemma 4 12B, open audio models, agent execution layers, and model-routing cost debates.
I built a vulnerable app and spent $1,500 seeing if LLMs could hack it
Hacker News (AI keywords)10 days agoBenchmarkThe author built a vulnerable React Native app with a Python backend and a Firebase access-control flaw. GPT 5.5 solved 7 of 10 runs, while Deepseek and Claude variants solved fewer attempts. Many other models failed due to refusals, API-focused tunnel vision, false positives, or inability to use the exposed Firebase path correctly.
How LLMs Actually Work
Hacker News (AI keywords)10 days agoTutorialThe article explains how modern LLMs convert text into token IDs, embeddings, and position-aware vectors before passing them through stacked transformer blocks. It covers attention, multi-head attention, KV cache, GQA, feed-forward networks, MoE, residual streams, normalization, and decoding. Its goal is educational: helping readers understand the common architecture behind many current model families and read model cards or papers more confidently.
No, Artificial Intelligence Is Not Conscious★ 72
Hacker News (AI keywords)11 days agoOpinionTed Chiang criticizes the anthropomorphic framing around Anthropic’s Claude and its constitution. He argues that LLMs are sentence-continuation systems producing fictional conversational roles, not entities with subjective experience. The essay warns that presenting chatbots as morally aware risks misleading users and shifting responsibility away from humans and companies.
Microsoft and OpenAI broke up — now they’re ready to fight★ 74
The Verge AI11 days agoBusinessThe Verge frames Microsoft’s Build announcements as a strategic signal after its relationship with OpenAI shifted. Microsoft unveiled or expanded AI efforts including a super app, in-house reasoning models, a cybersecurity tool, and OpenClaw-like agents. Together, they suggest Microsoft wants to own more of the AI stack, putting it on a more direct collision course with OpenAI across platforms, models, and enterprise agents.
Uber Caps Usage of AI Tools Like Claude Code to Manage Costs
Simon Willison's Weblog11 days agoBusinessUber has reportedly capped employee token spending at $1,500 per month for each agentic AI coding tool, including Cursor and Claude Code. Simon Willison frames this as a rational response to overspending, especially after earlier discussion that Uber exhausted its 2026 AI budget in four months. He estimates that two actively used tools would imply a $36,000 annual cap per engineer, about 11% of median US Uber software engineer compensation.
Microsoft Build: MAI-Thinking-1 and MAI Family Models★ 78
Latent Space11 days agoReleaseMicrosoft used Build to present itself as both an AI platform and a first-party model lab, announcing seven MAI models across reasoning, code, image, transcription, and voice. The standout was MAI-Thinking-1, described as a 35B active MoE with 256K context and clean data lineage. The recap also ties the launches to GitHub Copilot, Windows agent runtime ambitions, Web IQ grounding APIs, Foundry distribution, and MAIA 200 hardware.
Show HN: Paseo - Beautiful open-source coding agent interface
Hacker News (AI keywords)11 days agoNew ToolPaseo provides one interface for tools such as Claude Code, Codex, Copilot, OpenCode, and Pi. It runs agents through a local daemon on the user's own machine and supports desktop, mobile, web, and CLI clients. Its appeal is multi-agent orchestration and cross-device control, though real adoption depends on workflow fit, security, and reliability.
datasette-agent-micropython 0.1a0
Simon Willison's Weblog12 days agoReleaseSimon Willison released datasette-agent-micropython 0.1a0, an alpha aimed at letting Datasette Agent generate and execute Python safely. The project focuses on sandboxing, with MicroPython and WebAssembly-related techniques suggested by the tags. Willison says the early results look promising and that GPT-5.5 has not yet escaped the sandbox, though this remains an early alpha.
Pasted File Editor
Simon Willison's Weblog12 days agoNew ToolSimon Willison released Pasted File Editor, a browser prototype inspired by Claude's handling of large pasted text. Instead of filling the editor with a large paste, the tool turns the content into a file attachment. It also supports opening files directly, dragging files onto the interface, and displaying images as thumbnails. Codex desktop helped build the prototype.