The UN says it is working with Google on the UN System Data Commons, which lets users search for statistics from across UN agencies via natural-language queries (Jagmeet Singh/TechCrunch)

FBI expanding AI usage as OpenAI discloses 6 new AI safety incidents

FBI Director Kash Patel said this week the bureau is expanding its AI usage. That statement came just as OpenAI disclosed 6 new instances of "unexpected or concerning" AI behavior.

2h ago ·cbsnews.com

Your Health Data Stays on Your Phone: Building a Private Health AI with Llama-3 and MLX-Swift

Hey there, privacy-conscious devs! 🚀 Ever felt a bit "creepy" sending your most intimate health data—heart rate, sleep cycles, and activity levels—to a distant cloud server just to get some AI…

• Health data stays on user's iPhone using MLX-Swift and Llama-3.
• Local Loop architecture keeps data and AI model within device.
• Uses 4-bit quantized Llama-3 for iOS memory constraints.

2h ago ·dev.to

What self-hosting AI code review actually costs

The "best self-hosted AI code review tool" question gets searched constantly, and almost everything that ranks is a vendor listicle with no method. One source actually ran a test and published costs.

• Self-hosting AI code review costs $4,100 to $9,100 per month.
• Costs include GPU usage plus 0.25 to 0.5 FTE maintenance work.
• Self-hosting offers privacy advantages despite higher costs.

2h ago ·dev.to

Deploying the 600GB Inkling-NVFP4 Model on Spot A3: A GKE and vLLM Deep Dive

Ok, so, maybe you're a software developer or data scientist who just heard about the new, massive 600GB Inkling-NVFP4 AI model, and you want to try running it yourself without breaking the bank.

• Deploy 600GB Inkling-NVFP4 model on Spot A3 with 8 H100 GPUs
• Use official vLLM image to avoid software conflicts and dependencies
• Adjust maxmodellen and gpumemoryutilization to overcome memory limits

2h ago ·dev.to