does anyone practice dual build pipeline? eg: 1 by your devops team and another one by your security team and compare binaries hash later. To verify everything is reproducible.
If no one uses it, that means the market has proven, no audience for this kind of product. Google loses, everyone else loses.
If everyone who wants this sort of thing uses it, that's it, Google won, everyone else loses.
The outcome to sell to investors is the least believable: people will pay for some offering when a nearly identical one is available directly from Google for free. And anyway, they have the best generative creative tech, so how could anything be better than Google's?
The AI start-up field is going to be eviscerated. I too have the irresistible urge to bolt a SOTA LLM back end on a custom harness and charge $20/mo, but the total lack of a moat and ease of replication kills any motivation.
You have to either have some big cajones or be totally lost to think it's a good idea to create a startup that is just a simple cheap veil on someone else's extremely advanced and expensive product
For me Claude Skills is just a proof that we're making RAG unnecessary difficult to use. Not tech wise, but UX wise. But if we can fix that, the need for Claude Skills will go away.
Where Claude Skills is better than MCP? It's easier to produce a Claude SKills. It's just text. Everyone can write it. But it's dependant on the environment alot. Eg: when you need to have certain tools available for it to work. How do you automate sandbox setup with that? Even that, are you sure it's the right version for it to use, etc...
https://github.com/iced-rs/iced/releases/tag/0.14.0
reply