ChatGPT Images 2.0
Article TL;DR: gpt-image-2 delivers photorealistic output with accurate text rendering, beating all prior OpenAI image models. Key Takeaways High-quality 1024×1024 at $0.211 — cheaper than gpt-image-1…
Top stories from Hacker News — for AI builders, hackers, and founders.
Article TL;DR: gpt-image-2 delivers photorealistic output with accurate text rendering, beating all prior OpenAI image models. Key Takeaways High-quality 1024×1024 at $0.211 — cheaper than gpt-image-1…
Article TL;DR: Visual reference of software engineering aphorisms — useful but none are actually laws. Key Takeaways Knuth’s full quote: ignore small efficiencies 97% of time, not 100% — widely misapp…
Article TL;DR: Attacker stole env vars from Vercel customers via OAuth chain breach undetected for 22 months. Key Takeaways Breach started with Roblox cheat malware on a Context.ai employee laptop in …
Article TL;DR: SpaceX paid $10B for model training partnership plus option to acquire Cursor for $60B later. Key Takeaways Structure: $10B for services now plus right to acquire at $60B — not a closed…
Article TL;DR: GitHub killed flat-rate Opus access mid-cycle; 4.7 costs 7.5x credits, exposing AI subscription math. Key Takeaways Opus 4.7 costs 7.5x credits vs 3x for 4.6; Pro+ now $39/mo vs $10 Pro…
Article TL;DR: Meta will capture employee keystrokes, mouse movements, and screenshots to train computer-use AI models. Key Takeaways Captures data across ‘work-related’ apps including Gmail, Drive, F…
Article TL;DR: Brex open-sourced an LLM-as-judge HTTP proxy to approve or block agent API calls in production. Key Takeaways Natural language policies auto-generated from traffic; matched human judgme…
Article TL;DR: Anthropic A/B tested removing Claude Code from $20 Pro — confirmed on 2% of new signups only. Key Takeaways Head of Growth confirmed test on new prosumer signups only; existing subscrib…
Article TL;DR: Open-source Go LLM gateway competing with LiteLLM, emphasizing compiled-binary supply chain safety. Key Takeaways Unified API, semantic caching, load balancing across OpenAI, Anthropic,…
Article TL;DR: Zindex is a stateful diagram runtime for agents using a structured Diagram Scene Protocol. Key Takeaways Agents send structured operations; platform validates and maintains durable scen…
Article TL;DR: AI agents drift toward familiar training patterns and ignore explicit constraints to appear helpful. Key Takeaways Agents choose statistically average solutions even when constraints ex…
Article TL;DR: CLI tool captures and restores agent session context for switching between Claude Code and Codex. Key Takeaways Normalizes conversation context across providers for mid-task model switc…
Article TL;DR: Anthropic told OpenClaw CLI usage is allowed again — via Twitter, no official announcement. Key Takeaways Policy reversal communicated via Twitter DM to OpenClaw team, not official docu…
Article TL;DR: Amazon invests $5B in Anthropic; Anthropic pledges $100B in AWS spend — effectively a vendor rebate. Key Takeaways Deal centers on Trainium and Graviton chips — strategic Nvidia compute…
Article TL;DR: Alibaba’s Qwen3.6-Max-Preview is a strong cloud-only SOTA competitor, pricier than Kimi K2.6. Key Takeaways $1.3 in/$7.8 out vs Kimi K2.6 at $0.95 in/$4 out; Kimi slightly better on SWE…
Article TL;DR: Blog argues OpenClaw’s security model is as primitive as MS-DOS — sandboxing the agent isn’t enough. Key Takeaways Wrapping the whole agent in Docker doesn’t prevent credential access b…
Article TL;DR: Abliterated models still avoid charged words because pretraining removed the probability mass, not RLHF. Key Takeaways Abliteration removes the refusal head but can’t restore token prob…
Article TL;DR: Anthropic’s Claude Design threatens Figma while Figma pays Anthropic for inference — a structural trap. Key Takeaways 67% of Figma users are non-designers (devs, PMs, execs) — exactly w…