We've added a new blog post about the no-algorithm approach to be used by Narrative and how this is another example of member autonomy employed by the system:
Thanks for this update post - very informative.
It is encouraging that Narrative is taking such a distinctive stand around algorithmic targeting and related privacy concerns, but also that you seem already well advanced in articulating the advantage this confers to Narrative and its attractiveness.
As you know, a lot of niche owners are concerned they will not be able to make a compelling case to prospective high quality creators who already have rewarding outputs for their work, unless we can represent to them that their work will be featured in some way, at least for a certain period of time.
Reading this blog post, it occurred to me that a starting point for this sort of featuring could be developed within the framework you have put forth.
User sourced quality scores could make certain content qualify for inclusion in a 'featured content' area for each niche. A special type of niche owner's vote could carry more weight: enough to guarantee perhaps a 3 day featured status for content. In order to tailor this towards balance, you could limit the amount of such weighted votes an owner can make each week or each month. The same could be considered for moderators.
This still falls short of an ideal curation model, but I'm trying to come up with hybrid solutions that can work with your current direction. Ideally, owners and other curating moderators would be able to apply several levels of promotion to content. This would allow them to also promote excellent content by relatively unknown creators, giving them a moment in the spotlight and the chance for more eyeballs, leading to a higher quality score. Without this sort of ability, once the network is well underway, great creators who do not yet have a following will stand very little chance of getting noticed.
The current model is great in terms of giving users full control of their streams and protecting their privacy. That said, it strongly reinforces the echo chamber problem so many of us are sensitive to, and there is an easy and fun fix.
If you include a Surprise Me function, users could decide from time to time to be shown a collection of content from outside their self-created bubble. This collection could be populated in a hybrid manner, with a base of algorithm-selected content, but with human-curated content at the forefront.
The curated content should be selected with the challenge in mind: the only way someone will step outside their habitual content circle is if the new content is really valuable and readily identifiable as such. An example: I have no interest in Runway Fashion, but if a smart curator realises a certain piece of Runway Fashion content holds more universal appeal, like 'Three Reasons the Future of Runway Fashion Will Invade Public Spaces Near You' - I might actually click on that if only because the notion of runway fashion invading public spaces creeps me out! And so I get to read the weird and wacky story of how X plans to creatively do Y, with the help of Z, and all for a good cause (hopefully).
The algorithm-selected content could use similar criteria: analyse content that is getting higher user ratings from people who don't have any declared interests in its subject matter, and you will start isolating content that has the ability to transcend categories.
Everyone will benefit from being able to dip into a pool of this sort of content. But you still remain true to not imposing algorithmic selections on people based on their behaviours.
Also, we should not look at this as just a top down effort to try to free people from their echo chambers - this is also an extremely sensible business decision. We feel the most gratified by our routines when they bring us something new and refreshing in a context where we are receptive to it, and can turn it down easily if we want to. As long as such discoveries are opted into punctually, when the mood is right, the user will come away from that jaunt on Narrative feeling that with the Network, they have a window on a wider world with a guide to lead them to the most interesting spots, while still always remaining in control and able to retreat back to the familiar, on a whim.
Taking a simplistic approach for the upcoming Beta blogging software seems like a good idea. Similar to Medium, only showing bloggers what they follow is a safe start.
Malcolm's ideas on 'featured content' and 'surprise me' are good ideas. Perhaps another way to present featured content, would be to allow Narrators to pay to showcase their own work via token or fiat. This would allow burgeoning talent to showcase their topic pieces that might otherwise be lost in the digital wilderness. When we scroll through our Twitter timelines, this is in fact what we're doing when we read new RTs of non-followers. It's a form of non-paid advertising.
The number one reason why I disconnected from Fakebook years ago, was primarily the privacy intrusion, but also an endless stream of nonsensical bullshit was a huge time-waster. The FB blackhole was sucking up huge amounts of my time, and giving me nothing in return. I want a quality return on my time invested, and it simply wasn't delivering.
Malcolm also makes a good defense for algorithm curation taking people out of their echo-chambers. You know what would be unique? How about having a 'toggle switch' for taking people out of their comfort zones?
With the toggle off, people only see what they follow. With the toggle on, they enable algorithmic goodness to find them new and exciting content based on keywords in their own blogs or niches owned. (or something like that) That allows users that have a few extra hours on the weekend to turn their brain off and stroll through the community, and when those users are busy, they just turn off the suggestions.
Either way, looking forward to writing my first few pieces on the platform!