The results are strange but follow a particular formula (ask the model how it derived the result)
1. a mundane location, normal camera angle, ordinary props. (I think this is why bathrooms appear so often)
2. combining incompatible things (eg cereal + fish)
I think @Steve8708 is really on to something with agent-native. What if you could start building any react app that immediately offered agents to your users: skills, context, eyes, subagents, agent control + communication with the UI... all for free. agent-native.com
toddmorey retweeted
We made builds on Netlify a lot faster.
Over the last year we've been building our own MicroVM based compute platform.
It first went into production as our sandbox infrastructure for Agent Runners.
Now it's live for our build system, and the effect is dramatic.
toddmorey retweeted
@huggingface + Label Studio SDK = 🔥
• Easily import HF datasets into Label Studio for annotation
• Connect HF models with Label Studio for evals, pre-annotations and active learning
• Export labeled data for model training & fine-tuning
Tutorial ↓ labelstud.io/tutorials/how_t…
toddmorey retweeted
We’re big fans and supporters of @tan_stack and it’s now default when starting new projects on Netlify.
Really excited to have @jherr on our team!
Looking forward to both his continued contributions to TanStack and the work he’ll do to make Netlify the best place for builders.
Welcome aboard, @jherr!
If you know Jack, you know he’s been deep in @tan_stack AI, MCP, and the tooling behind AI apps that actually ship.
toddmorey retweeted
Has anyone built an AI app that ingests your health insurance plan, audits all your medical & claims activity, makes sure Aetna or whoever are paying/reimbursing what you are entitled, keeps track of deductibles, disputes claims on your behalf, etc? This is still an unsolved problem, feels totally tractable at this point
