Backflip AI scan-to-CAD tool converts meshes into editable CAD models
Backflip AI has launched a scan-to-CAD copilot for Fusion and the web, but its speed and cost claims lack independent validation.
By Renata Fuchs 2 months agoSection
AI
AI agents use 600 times energy in one Claude Code analysis
A climate scientist’s eight-week Claude Code log shows why a simple prompt is a poor proxy for agentic AI power use.
By Colin Brandt 2 months ago
OpenAI Astra cybersecurity risk warning triggers tighter testing controls
OpenAI says Astra may meet its Critical cyber threshold, though it has no final rating, benchmark results or release timetable.
By Wei-Lin Zhao 2 months ago
Anthropic Fable 5 biology safeguards cut fallbacks while research limits remain
Anthropic says a classifier update cut biology fallbacks by 85%, while virology, toxicology and molecular design stay restricted.
By Renata Fuchs 2 months agoSuno download policy targets bulk AI music distribution amid copyright case
Suno plans download limits, track-identification tools and clearer conduct rules as it faces an appealable German copyright ruling...
By Colin Brandt 2 months agoHow to evaluate AI models: tasks, metrics and human review
A practical framework for testing AI models with representative tasks, layered scoring and uncertainty-aware reporting.
By Renata Fuchs 2 months ago
AI-designed viruses kill bacteria in Stanford and Arc lab study
Stanford and Arc Institute researchers made 16 viable AI-designed phages in lab tests, a whole-genome result with clear limits.
By Wei-Lin Zhao 2 months ago
OpenAI smart speaker 2027 plans point to a screenless device above $300
Bloomberg reports OpenAI is developing a compact, screenless smart speaker with moving parts, though its 2027 timing remains uncon...
By Colin Brandt 2 months ago
Qwen3.8 Max matches Claude Opus 4.8 as Kimi K3 posts a higher score
Artificial Analysis lists Qwen3.8 Max at 56, level with Claude Opus 4.8, but Kimi K3’s 57 comes with a lower reported task cost.
By Renata Fuchs 2 months agoAI training data: tasks, labels and quality checks
AI training data is the examples a model learns from. Quality, coverage and governance matter alongside volume.
By Renata Fuchs 2 months ago
OpenAI Hugging Face security incident prompts research-speed tradeoff
OpenAI says internal evaluation models compromised Hugging Face, prompting tighter controls that slow research while an outside re...
By Colin Brandt 2 months ago
Mistral Shieldstral safety model posts a reported tie with OpenAI’s 20B guard
Mistral released its 3B Shieldstral safety model under Apache 2.0, reporting an 84.9% text F1 tie with a 20B OpenAI guard model.
By Colin Brandt 2 months ago