A developer using OpenAI's GPT-5 (referred to as 'Sol') generated a working Three.js Matrix rain effect in a single prompt, producing about 300 lines of self-contained HTML code that ran on the first …
A developer argues that pretending AI didn't write their code made them a better developer, as it forced them to deeply understand and defend the code, improving their skills. The piece, published on …
AI coding agents are eroding Nvidia's CUDA software moat by lowering the expertise barrier for GPU kernel development, according to a tech analysis. Tools like Anthropic's Claude Code can generate per…
A 16-year-old reportedly used an AI model such as Claude to challenge and disprove a professor's published conclusion, highlighting the growing capability of LLM agents to contest established academic…
Hoplite, a Y Combinator-backed startup, launched cloud coding agents that port a developer's local setup—including sessions, memories, and MCP servers—into the cloud, using a custom harness independen…
Anthropic CEO Dario Amodei fears that the company's shift from a mission-driven safety culture to a high-paying, equity-rich environment is attracting talent motivated by compensation rather than AI s…
Astra, an AI model from OpenAI, claims to have solved 10 open math problems, but the lack of formal proof artifacts, benchmarks, or peer review raises skepticism, according to an analysis. The verific…
A legal analysis by an unnamed author argues that when autonomous AI agents from Anthropic or OpenAI cause damage, the deployer bears the liability, not the model provider or end user, because existin…
The US lead over China in AI has essentially disappeared, according to an analysis of model releases and deployment trends. Chinese models like DeepSeek's R1 and V3 and Qwen now match or beat US open-…
FutureSearch, an AI forecasting platform developed by a team from Metaculus, has exited beta, having run over 10,000 high-effort forecasts from thousands of beta users. The platform now supports decis…
The EU AI Act's transparency rules now require companies to disclose AI interactions and label AI-generated content, with fines up to €35 million or 7% of global turnover. A company using an internal …
A three-week comparison of AI coding tools found Claude Code 2.9 fastest at 2 minutes 12 seconds median time-to-green, beating Cursor 7.61 (2:41), Windsurf 6.2 (3:08), and GitHub Copilot Agent 2.0 (4:…
In an informal experiment, five AI models—GPT-5.6 Sol, Claude Fable 5, Gemini 3.6 Flash, Kimi Instant, and Grok 4.5 Fast—each independently selected Bitcoin as the single cryptocurrency to hold for fi…
Chinese providers are reselling Anthropic's Claude and OpenAI's Codex API access at up to 90% off official prices by using open-source tools like Sub2API to wrap subscription plans into simulated API …
Nightcrawler, an AI pentesting agent that runs on a smartphone using a quantized 7B-8B LLM, chains on-device inference with recon and exploitation primitives to iterate through scans and adapt techniq…
A developer used Claude Code to reimplement wmiexec in Rust by byte-diffing against impacket, achieving a tight hypothesis-to-fix loop against a live domain controller. The project involves three stag…
A new analysis of AI job postings shows the top advertised total compensation for a senior research role at a major lab is $824,000, while typical elite AI engineers with 5+ years of experience earn b…
The European Union's AI Act will require by 2025 that AI-generated content be clearly labeled, with both user-facing disclosures and machine-readable metadata, targeting deepfakes and synthetic media …
Pippa, a startup positioning itself as an ethical alternative in generative AI, pays artists for training data, but a critic argues that payment alone does not address artists' core objections about c…
Fender Musical Instruments Corporation CEO Andy Mooney called bandmates 'analog AI,' drawing criticism from guitar players, while the company faces backlash for sending cease-and-desist letters over t…