GPT-5.6 used prompting to bridge a 30-year gap in convex optimization research.
// curated from Hacker News with AI
GPT-5.6 used prompting to bridge a 30-year gap in convex optimization research.
NYC may require landlords to disclose AI use in property listings to prevent deception and protect tenants.
AI logos often resemble buttholes due to circular design, industry conformity, and risk aversion, highlighting sameness in tech branding.
AI impacted Stack Overflow's data access, prompting security checks to verify human interaction.
Fable 5 outperformed GPT-5.6 on an NP-hard problem; /goal mode sometimes helps, but persistence matters more than the feature.
Convert a spare Mac into a controlled environment for Claude Code via SSH, enabling remote access, automation, and secure task delegation.
OpenAI resets Codex usage limits frequently during rapid growth, system scaling, and incidents to ensure users can innovate seamlessly.
Google DeepMind and Isomorphic Labs collaborate on bioresilience, advancing partnerships, safe AI models, outbreak detection, and rapid vaccine development.
Digital evolution shows self-replication and problem-solving co-evolving, driven by environmental tasks shaping reproductive and functional complexity.
AI hype compromises decision-making, exposes failures, fosters false claims, and disconnects organizations from reality.
AI's integration into software engineering turns coding into tedious, slop-like work, diminishing craft and enjoyment for developers.
Anthropic's unsettling ad faced backlash for dark imagery, evoking paranoia and criticism, risking brand reputation in AI ethics debate.
Claude Fable 5 added to all Max plans from July 20, with limited use; standard users get credits and a $100 bonus.
AI hype is causing global decision paralysis; most AI investments fail, fostering a culture of dishonesty and risk aversion in organizations.
History repeats as AI bubble mirrors the Dot-Com crash with overvaluations, insider sales, and speculative trading.
Gwern suggests overtraining large models on small data may induce human-like understanding, potentially unleashing true AI generalization.
AI employees outspend and coordinate more in politics than previous tech cohorts, shaping influence, especially in San Francisco.
Anthropic is in early talks to lease Meta’s computing power, possibly a $10B deal, as both expand AI infrastructure collaborations.
David Gerard criticizes flawed, poorly written AI code, calling it "insane trash" in a video critique.
Open Design, an open-source AI tool, gained 77k GitHub stars and 1M+ installs in just 11 days by leveraging human-agent collaboration and file-based contracts.
AI subscription prices, anchored at $20 by OpenAI, are industry standards influenced by market perception rather than cost.
AI development remains bottlenecked by verification and runtime challenges, not code review, in cloud-native and infrastructure contexts.
PixelUp offers offline, lightweight AI video upscaling for Windows, enhancing videos up to 8X with privacy and no subscriptions.
Claude Fable 5 stays permanent for Max and Team plans, ending the threat of its removal amid competition from GPT-5.6.
Transforms Ilya Sutskever's key AI papers into an interactive RPG, making mastering 90% of critical AI knowledge engaging and gamified.
Google workers petition CEO for layoff protections amid AI-driven workforce reductions, expressing concerns about job security and profit priorities.
AI pipeline automates cycling highlight reels by ranking footage using visual interest, telemetry, and learned taste; built for personal passion.
AI detection resists evasion attempts, even with multiple AI-guided edits and voice mimicry, consistently flagging AI-generated content.
Trump administration now controls access to frontier AI models, prioritizing security and government approval over private company decisions.
Peek-CLI captures browser screenshots for AI agents, enabling iterative front-end design testing with Claude Code and others.
Anthropic claims their language models have a global workspace, but measurement issues and unfalsifiability cast doubt on true AI consciousness.