We are still a very long way from distributed discovery.
Back in the day at Demon Internet (well known 90s UK ISP)
every customer got a free 10MB of webspace - barely any of them used it.
But it should have been how we all hosted our blogs and our videos. And it would solve a multitude of social media issues.
If Facebook and youtube did not host its hard to see how they could compile trustable usage stats (#) and so hard to see how we could all find out we must watch Gangnam Style. Similar problems seem to hover over crypto-currencies.
Edit: some random thoughts
- DNS: In a world of individual producers, curated aggregation still has value as it provides a quality / interest gate. A TV channel implies curation, a review guide the same. Perhaps it is no coincidence Marvel has found its footing in a world where choice is infinite and curation zero. Every TV show hopes to land a few million dedicated fans.
- Social media companies basically have zero curation (this is weirdly tending towards including Amazon who are dumbing down curation in marketplace)
- Our natural curation signals are usually based around proximity (geography, recommendation, similarity). These fall apart where the same DNS hosts nazis and nannies. Especially where recommendations are controlled by (one?) algorithm.
- If enough viewing moves off YouTube (not merely the bits over pipes, it off the influence of the algorithm) then these other effects might start to show through.
- But even if we could gather reliable distributed usage stats (book publishers managed it for decades), distributed discovery becomes a strange animal. A dumb algorithm that recommends what is the most watched video in the world, followed by the second, would have some strange chaotic behaviour but would at least have twin virtues of simple and transparent. Start adding in anything else (here is my past history, make a recommendation, or here is my aspirational set of people (my twitter follow list is watching) or just show me what Barry Norman recommends, all of these can be made transparent - and open to a cottage industry of discovery algorithms.
- This cottage industry fascinates me - it's like the other industries I expected to exist but don't (decent job search, dis-intermediated real estate). Duck Duck Go seems the model here - there is likely to be a (regulated?) split in discovery - much like railways or phone lines - where the underlying monopoly bit (we scraping) gets hived off and everyone can access it at common fees - and the add on bits, the start of distributed discovery - appears as a more competitive market. And one hopes a less behaviour driven one (I am still wanting to know the revenue difference between Google storing my every move and returning what it thinks I want and Duck Duck Go just using the context of my query - and returning what I asked for.
- But hopefully after a Cambrian explosion of companies offering discovery - not just search or for videos and entertainment but even shopping (as Amazon hits the same split the Google inevitably will), after all that, we are still a long way from where I hope we can get.
- Everything online should treat me as a citizen, as even a patient - where do no harm is the first principle. If we live in an internet where my individual best interests are the legal and cultural norm, as professions are supposed to be, then we enter a totally different equation. We are all exploited online now, and the discussions are about harm reduction. But that's not the real end goal. Making sure quacks don't charge too much for the snake oil is not what made medicine work.
(#) ignoring facebook lying to advertisers etc etc