bnolan/gist:61b5c90860ef9686e6470734fefbb529

## gistfile1.txt
i have a great idea for a super ambitious project
8:07
its a search engine that has only a voice interface
8:07
you specify a little subset of html, and get people to publish their content in that format. the html is marked up with speech to text hints so that you can read it out with emphasis.
8:08
they then submit a link to their content to your search engine, and you index it and let people search it with voice.
8:08
the actual voice search you use the on-device apis provided by google or apple, you dont actually have to do that yourself.
8:08
then once that’s working, you tell people on hacker news, and they generate lots of random bits of content for your search engine to gobble up.
8:09
its like creating an alexa task, but way easier, since you just have to make an html file and add it to the index.
8:09
then, your app that does the recognition, you ask it a question, and then it fetches the content from the publishers site, and then reads out the answer with speech to text
8:10
so content publishers can see how many people (or lack of) are visiting their content.
8:10
you also add support for some very basic ‘form’ markup, that lets people create old school html forms that get read out
8:10
its a combination of a screen reader / web browser / search engine
8:11
anyway, that’d be a super amibitous project. the cool thing would be to make it really easy to submit your content for the index, so that people could quickly add searchable content to the index.
8:11
that’s where you win, because people are custom making content for your app, because it’s quicker and easier than creating an alexa skill, or siri skill or google voice content.
	i have a great idea for a super ambitious project
	8:07
	its a search engine that has only a voice interface
	8:07
	you specify a little subset of html, and get people to publish their content in that format. the html is marked up with speech to text hints so that you can read it out with emphasis.
	8:08
	they then submit a link to their content to your search engine, and you index it and let people search it with voice.
	8:08
	the actual voice search you use the on-device apis provided by google or apple, you dont actually have to do that yourself.
	8:08
	then once that’s working, you tell people on hacker news, and they generate lots of random bits of content for your search engine to gobble up.
	8:09
	its like creating an alexa task, but way easier, since you just have to make an html file and add it to the index.
	8:09
	then, your app that does the recognition, you ask it a question, and then it fetches the content from the publishers site, and then reads out the answer with speech to text
	8:10
	so content publishers can see how many people (or lack of) are visiting their content.
	8:10
	you also add support for some very basic ‘form’ markup, that lets people create old school html forms that get read out
	8:10
	its a combination of a screen reader / web browser / search engine
	8:11
	anyway, that’d be a super amibitous project. the cool thing would be to make it really easy to submit your content for the index, so that people could quickly add searchable content to the index.
	8:11
	that’s where you win, because people are custom making content for your app, because it’s quicker and easier than creating an alexa skill, or siri skill or google voice content.