Polywork is not planning to have "likes/reactions" and it had me wondering recently how that affects discoverability or how would one engineer a system that helps with fair or good discoverbility?
I just posted an article the other day on DEV which I have not done in a while, and my primary goal was to get enough reactions within the first hour to get trending so I had any chance of people seeing the article, while it did not reach that goal, people were saying they read it on twitter (instead of just likes) and so that was my metric of impact.
I suppose search engines have been doing it (discoverability) for a long time, though I'm guessing for social platform you'd want to use a classification ML model to say this article goes to these tags, and you could build an ML model to determine the quality to help give articles a boost and then let commentors or page interactions be the defining metric.
Human-assistance is an option for good article curration and I did about 30 mins of DEV moderation for AWS tag yesterday.I just got tired and was thinking, an AI-assistant could cut this down to 90% so the human can focus on uplifting the good stuff.
I'd like to see DEV just get rid of reactions, or at least have them not determine the timefeed. I think this could help cut down on the buzzfeed list-like content.
I really like reading your articles Helen, especailly since right now I've been diving so much into data business intelligence tools. Hope to see you at re:Invent permitting the worlds current circuumstances and mother gia allows.
👋 Hey there, I am Waylon Walker
I am a Husband, Father of two beautiful children, Senior Python Developer currently working in the Data Engineering platform space. I am a continuous learner, and sha
Polywork is not planning to have "likes/reactions" and it had me wondering recently how that affects discoverability or how would one engineer a system that helps with fair or good discoverbility?
I just posted an article the other day on DEV which I have not done in a while, and my primary goal was to get enough reactions within the first hour to get trending so I had any chance of people seeing the article, while it did not reach that goal, people were saying they read it on twitter (instead of just likes) and so that was my metric of impact.
I suppose search engines have been doing it (discoverability) for a long time, though I'm guessing for social platform you'd want to use a classification ML model to say this article goes to these tags, and you could build an ML model to determine the quality to help give articles a boost and then let commentors or page interactions be the defining metric.
Human-assistance is an option for good article curration and I did about 30 mins of DEV moderation for AWS tag yesterday.I just got tired and was thinking, an AI-assistant could cut this down to 90% so the human can focus on uplifting the good stuff.
I'd like to see DEV just get rid of reactions, or at least have them not determine the timefeed. I think this could help cut down on the buzzfeed list-like content.
I really like reading your articles Helen, especailly since right now I've been diving so much into data business intelligence tools. Hope to see you at re:Invent permitting the worlds current circuumstances and mother gia allows.
Never heard of polywork. I just signed up and waiting to get in.
Thanks for your post on Polywork Andrew. I'm also interested to see how the platform is received.
Is PolyWork the new LinkedIn for Developers?
Andrew Brown 🇨🇦 ・ Jun 17 ・ 6 min read
I'm crossing my fingers that the world is not on fire in December and we get to re:Invent. This may be the year we meet in person :D