tl;dr: This is a link to a sign up sheet for a yet-to-be delivered product promising something that is vaguely hinted at - "Drive traffic [...] AI-powered [...] help you create and share content [...]" - nothing to see here.
And yet the CTA touches on something I find deeply interesting: how can we use LLMs to make products more discoverable to people using LLMs to search for products to solve their particular problems.
I put some work into this over the summer - prepping my canvas library repo (and my author site, but that's just vanity) with LLM-friendly documentation and training data which I hope will, at some point, get picked up by the AI data scrapers and crawlers for inclusion in the next generation of training data, to be used in turn to train the next generation of LLMs.
The big downside to this approach is, of course, the terrible lead-in times (6-18 months) alongside the lack of guarantee that the docs I've written will be good enough to pass muster and make it into the training data sets.
But the payoff I'm after is to get accurate, unbiased information about my canvas library into LLMs in such a way that when people ask general questions about coding up the canvas element my library gets ranked alongside the usual suspects, and gets assessed fairly against them. Plus having accurate "how to" templates to minimise the glaring mistakes current LLMs make when hallucinating how they think my library should be used.
rikroots•10h ago
And yet the CTA touches on something I find deeply interesting: how can we use LLMs to make products more discoverable to people using LLMs to search for products to solve their particular problems.
I put some work into this over the summer - prepping my canvas library repo (and my author site, but that's just vanity) with LLM-friendly documentation and training data which I hope will, at some point, get picked up by the AI data scrapers and crawlers for inclusion in the next generation of training data, to be used in turn to train the next generation of LLMs.
The big downside to this approach is, of course, the terrible lead-in times (6-18 months) alongside the lack of guarantee that the docs I've written will be good enough to pass muster and make it into the training data sets.
But the payoff I'm after is to get accurate, unbiased information about my canvas library into LLMs in such a way that when people ask general questions about coding up the canvas element my library gets ranked alongside the usual suspects, and gets assessed fairly against them. Plus having accurate "how to" templates to minimise the glaring mistakes current LLMs make when hallucinating how they think my library should be used.