This weekend brought us both a big New York Times piece titled “The Internet is Broken: @ev​ Is Trying to Salvage It” and a Guardian article about Facebook’s guidelines for content moderators. Both speak to the question of who controls content on the Internet and how that control impacts our attention. Ev makes the point that extremes attract more attention compared to a more moderate or balance view. Therefore advertising based systems, which require attention which they then partially resell (that’s the fundamental nature of advertising), are unlikely to ever be the best guardians of human attention. One alternative are subscriptions, but for that to really work for news, commentary, or even just friends’ status updates we need to separate aggregation/discovery from publishing. 

One of the great breakthroughs of the web was that it brought us permissionless publishing. We no longer required a gatekeeper publisher but instead could each have our own website or blog. And initially that’s how the web grew. Millions of individual sites. But there was a content discovery problem in this fragmented world. This initially gave rise to aggregators. Eventually though, a new model emerged that re-centralized the web by recombining aggregation/discovery with publishing. Ev was in fact the founder of three such integrated platforms: Blogger, Twitter and Medium. Facebook, the largest platform on the web today is of course another example that combines these functions (and at USV we have been investors in Twitter and Tumblr which both did this).

Why keep aggregation/discovery and publishing separate? For two reasons: first, it does not provide a central point for censorship. Second, it allows for aggregators/discovery solutions to compete with each other. In such a world we would not need to debate whether Facebook’s policies are the right ones. Instead, all content could get published and you can choose an aggregator that best reflects your own needs and values. 

It is possible to make a subscription model work in that world also. Blendle and Scroll are both working on that for larger publishers. Patreon is an alternative subscription model focused on directly subsidizing longtail content creation. I would also happily subscribe to Techmeme, if Techmeme kept only a small percentage of that for its aggregation services and passed the rest on to the content creators (on the basis of which content I click through from Techmeme). Same goes for Nuzzel.

Many of the investments we have made recently at USV are aimed at building a decentralized infrastructure for publishing and aggregation/discovery. For instance, Brad recently announced our investment in Protocol Labs which has developed the IPFS protocol and is now working on Filecoin. And our portfolio company Blockstack has developed a namespace and storage system that can be used to provide identity in a decentralized publishing world.

So yes, the Internet may be broken, but the way to “salvage” it is undo the bundling of publishing and aggregation/discovery through decentralized systems.

This weekend brought us both a big New York Times piece titled “The Internet is Broken: @ev​ Is Trying to Salvage It” and a Guardian article about Facebook’s guidelines for content moderators. Both speak to the question of who controls content on the Internet and how that control impacts our attention. Ev makes the point that extremes attract more attention compared to a more moderate or balance view. Therefore advertising based systems, which require attention which they then partially resell (that’s the fundamental nature of advertising), are unlikely to ever be the best guardians of human attention. One alternative are subscriptions, but for that to really work for news, commentary, or even just friends’ status updates we need to separate aggregation/discovery from publishing. 

One of the great breakthroughs of the web was that it brought us permissionless publishing. We no longer required a gatekeeper publisher but instead could each have our own website or blog. And initially that’s how the web grew. Millions of individual sites. But there was a content discovery problem in this fragmented world. This initially gave rise to aggregators. Eventually though, a new model emerged that re-centralized the web by recombining aggregation/discovery with publishing. Ev was in fact the founder of three such integrated platforms: Blogger, Twitter and Medium. Facebook, the largest platform on the web today is of course another example that combines these functions (and at USV we have been investors in Twitter and Tumblr which both did this).

Why keep aggregation/discovery and publishing separate? For two reasons: first, it does not provide a central point for censorship. Second, it allows for aggregators/discovery solutions to compete with each other. In such a world we would not need to debate whether Facebook’s policies are the right ones. Instead, all content could get published and you can choose an aggregator that best reflects your own needs and values. 

It is possible to make a subscription model work in that world also. Blendle and Scroll are both working on that for larger publishers. Patreon is an alternative subscription model focused on directly subsidizing longtail content creation. I would also happily subscribe to Techmeme, if Techmeme kept only a small percentage of that for its aggregation services and passed the rest on to the content creators (on the basis of which content I click through from Techmeme). Same goes for Nuzzel.

Many of the investments we have made recently at USV are aimed at building a decentralized infrastructure for publishing and aggregation/discovery. For instance, Brad recently announced our investment in Protocol Labs which has developed the IPFS protocol and is now working on Filecoin. And our portfolio company Blockstack has developed a namespace and storage system that can be used to provide identity in a decentralized publishing world.

So yes, the Internet may be broken, but the way to “salvage” it is undo the bundling of publishing and aggregation/discovery through decentralized systems.