Need to keep up with the news? Searching for a recent article on gravitational waves? Bored at work? Wondering if that 12 seed is gonna pull off the upset? Say hello to DiggBot, launching today on Slack.
01110100 01101000 01100001 01101110 01101011 00100000 01111001 01101111 01110101 00100001
Why Build A Bot?
Digg’s move into bot-land starts with a platform bet, that leads to an affinity for AI Overlords (we love AlphaGo). Here’s an observation: the messaging market is huge. There were 1.4 billion monthly active users on messaging apps by the end of 2015 (over 3 billion if you don’t de-dupe), sending over 90 trillion messages last year, and that number in aggregate is growing faster than any single network in the world. Our daily attention is increasingly captive within messaging, and we have noticed more and more users sharing links they’ve found on Digg directly into these apps. But there still isn’t an easy way to discover great content directly through messaging.
Of course, content discovery has to look very different in an app built for text conversations (duh). Which is why a bot makes the most sense. At its best, a bot will help us find the best articles and videos through conversation, and sometimes send us great stuff without needing to be told. That way, we can discover the best of the web, share it, and talk about it all in one place.
(Btw, if you want to learn more about bots, read this awesome article from The Verge, then this one from Motherboard, then try this one from Monday Note, and end on this post about a bot our betaworks family has built.)
Big Data, Digg Data
So what exactly makes DiggBot so cool and smart? Data. Every day, Digg aggregates 10 million RSS feeds from virtually all publishers. We collect 200 million tweets daily, 40 million of which include links. From there, our algorithms process all of these links in real-time.
The result: 7.5 million articles and videos collected, indexed, clustered, and ranked … every single day.
And on top of all this data, we’ve also got an incredible team of human editors cherry-picking the best stories and videos and training our algorithms to recognize great links.
In the end, you have all the content on the web, ranked by the best quality filters, in one place. Yeah, we’re excited too.
What Can DiggBot v1 Do?
Our bot is still a newborn, but it’s already learned some pretty cool tricks. It can find the best and most timely content based on keywords, and will also send you stuff throughout the day that is trending or important. And if you’ve been buried in an Excel model since dawn, don’t worry: twice a day, DiggBot will send you a package of articles and videos to catch you up. DiggBot is still a little shy, and we’re still teaching it to talk. But give it a few weeks, and we’re sure it’ll become friendlier. Also, we’re letting DiggBot play in Slack first, but it’ll pop into other messaging playgrounds soon.
Oh and did we mention that DiggBot ❤ March Madness?
Here’s a list of commands DiggBot understands:
Gives you live 🏀 scores and news from the NCAA Tournament
FOR A LIMITED TIME/digg
Gets a cool, fresh link. Simple./digg (keyword)
Search content by keyword/digg (domain)
See top trending stories for a domain (e.g. “/digg nytimes.com”)/digg trending
Get a list of stories that are trending on the internet/digg trending (on/off)
Turn trending alerts on and off (about 3–5 a day)/digg edition
See the most recent Digg Edition, a curated package of news, videos, and interesting stories./digg edition (on/off)
Turn Daily Edition delivery on and off (once a day, in the AM)/digg fun
Gets you something… fun