Post

Chrome looks set to ship an LLM Prompt API to the web platform. At Mozilla, we oppose this API. We feel it has a large interoperability risk, and Google imposing T&Cs on a web API sets a dangerous precedent. Full details: github.com/mozilla/stan...

... and we already have the fetch API, which solves this very well today.

Exactly what I was going to say. LLMs are just one more service you might want to call. And just like any external service, you need exercise extreme caution in what you let it do to your JavaScript and DOM.

The difference here between fetch and the prompt API is that the Prompt API is running client side with a pre-fetched model. Big difference.

Provide an alternative proposal.

300%. Endless naysaying with no fucking plan, no assistance, just endless little shitty nattering crappy demands. I'm fucking pissed what a savage hack job Mozilla has been against the web for too long, savaging very fine attempts to improve the web. I trust Jake much more but still pissed.

they aren’t obligated to..?

Yeah, I believe so, if they are fighting the same fight (whether the web will be relevant at all on desktop because of AI)

No they aren't! And I as a web developer am not obligated to support their browser.

The app maker can pay for their own tokens on the backend, like they pay for their own VMs, storage, network routing, software development, and debugging. They need to anyway, because an api that provides some random LLM at the user’s cost, or sometimes no LLM, can‘t actually be used predictably.

Ew. I dont trust or want app makers running their models. I don't trust them at all. I want my user agent giving me local & other models that I control. I absolutely detest the idea of having my sessions go to the app: that data is far too private, & longer term I can't trust their model training.