Wait is Cursor selling licenses to cities now?

Wow this is such a brilliant idea for running AI models locally. 🎯 webFrame is webAI 's backend that slices a huge language model into smaller shards, sends each shard to a different computer on your own network, then stitches the answers back together on the fly. Because

Hire the engineer who worked on the best product Hire the marketer and sales person who managed to get the worst product to outsell the best product

Privacy by policy fails. Current AI lawsuits exposes it: user chats, even deleted ones, ordered for discovery despite “privacy” claims. Only privacy by design—data staying on your device protects you.

webAI Knowledge modeling is next level. Same document, same system prompt. Also runs on local hardware.

models continue to race to bottom ... this is good for the world -- thanks OpenAI Models are free.

build build build

Anyone else think these OSS OpenAI models are trash? Feels like they intentionally nuked it with the quantization .

GPT-5 isn’t what we expected. On public benchmarks, 25–35% of failures weren’t random — they came from ambiguous, context-heavy tasks that monolithic models just can’t nail. This vision of “one giant brain for everything” is ending. The future? Networks of highly contextual

On the latest TinyLlama 1.1B benchmarks (MacBook Pro M4): Unquantized: ~8× faster than llama.cpp 8-bit & 4-bit quantized: still leading — and getting faster month over month From July → August, every variant of webframe-cpp improved its speed.

Morning. On today’s show: – Delian and Everett debate – GPT-5, one week later – China’s AI training woes – Should Trump buy Intel? Guest call-ins: – delian & Everett Randle (Founders Fund / Kleiner Perkins) – Bill Bishop (Sinocism) – Lennart Heim (RAND) – David Stout (webAI) –

We asked David Stout (Co-Founder & CEO, webAI) about the advantages of running AI on devices. "There's no free lunch, but there are things cloud players can't do. It's RAM intensive, yes, but by distributing across devices, we can unlock more accuracy and context. But if you're