An honest braindump on how pairing Gemini Antigravity (Gemini 3.8 Flash on Medium effort) with Cloudflare eliminated site maintenance friction, accelerated content, and transformed both my blog and my wife's clinic website.
Everyone is complaining about low-quality AI code and 'vibe coding'. But code quality didn't crash because of AI. It crashed because humans are rubber-stamping code they didn't review.
After removing image backgrounds in the browser with WebAssembly, I wanted to see what else we can run locally. Here are live interactive demos for on-device LLM text generation and voice synthesis.
My earlier in-browser AI post tested lightweight 135M–500M models. But what if you want a genuinely smart, GPT-3.5-caliber model like Llama 3.2 running on your device? Apparently, you can run much bigger models if you use WebLLM.
Back in 2012 and 2013, I used hacky ImageMagick bash scripts to isolate image backgrounds. Here is how modern background removal works today with real before and after comparisons using AI models, rembg, WebGPU, and native OS tools.
