I'm with you most of the way, Justin. The exception is your argument that "we should want companies which use technology, including algorithms, to promote positive and diverse conversation and ideologies". Allow me to repeat what I wrote in my own post on "Why I'm staying with Substack": "Perverse actors will always try but the enticemen…
I'm with you most of the way, Justin. The exception is your argument that "we should want companies which use technology, including algorithms, to promote positive and diverse conversation and ideologies". Allow me to repeat what I wrote in my own post on "Why I'm staying with Substack": "Perverse actors will always try but the enticement, for them as for legitimate writers, stems from Substack’s recent efforts to imitate social media by promoting/pushing other sites on the platform. I thought that getting away from all this pushing and algorithm-based manipulation was part of the point of writing and reading on Substack!"
My advice is to "Develop the Explore function further and let visitors/writers find their own preferred sites! And don’t be greedy."
At this point, Substack owners appear far from accepting my advice. Most of the emails I get from platform management are about "growing." Some of that is fine, of course, but the Recommendations algorithms should be abandoned. The bad that comes from it is outweighted by the "little blades of grass" that poked through in your experiment.
I've also long believed, like reader MacDowell, that anonymity does not encourage civil behaviour. I understand the need for anonymity in certain circumstances but that could be accommodated by special permission from the platform.
I think those are some pretty good points. But I also don't think that all algorithims are created equal. There's one version, yes, that tries to juice a user's preferences to get them locked in a cycle of consumption, regardless of the impact (Youtube used to be like this, Tiktok is like this now.) But algorithims can also be much gentler, built towards nudging you towards things you might not know you like. We're right to be skeptical of everything, here, but what I've seen thus far bodes well.
I'm with you most of the way, Justin. The exception is your argument that "we should want companies which use technology, including algorithms, to promote positive and diverse conversation and ideologies". Allow me to repeat what I wrote in my own post on "Why I'm staying with Substack": "Perverse actors will always try but the enticement, for them as for legitimate writers, stems from Substack’s recent efforts to imitate social media by promoting/pushing other sites on the platform. I thought that getting away from all this pushing and algorithm-based manipulation was part of the point of writing and reading on Substack!"
My advice is to "Develop the Explore function further and let visitors/writers find their own preferred sites! And don’t be greedy."
At this point, Substack owners appear far from accepting my advice. Most of the emails I get from platform management are about "growing." Some of that is fine, of course, but the Recommendations algorithms should be abandoned. The bad that comes from it is outweighted by the "little blades of grass" that poked through in your experiment.
I've also long believed, like reader MacDowell, that anonymity does not encourage civil behaviour. I understand the need for anonymity in certain circumstances but that could be accommodated by special permission from the platform.
Keep up the good work.
I think those are some pretty good points. But I also don't think that all algorithims are created equal. There's one version, yes, that tries to juice a user's preferences to get them locked in a cycle of consumption, regardless of the impact (Youtube used to be like this, Tiktok is like this now.) But algorithims can also be much gentler, built towards nudging you towards things you might not know you like. We're right to be skeptical of everything, here, but what I've seen thus far bodes well.