Social Search Engines - SSE
There are so many data nowadays that it is becoming harder than ever to find interesting ideas. A needle in a haystack comes to mind. The other idea that comes to mind is: but we have Google!... And it is true that PageRank is an idea to help you to find relevant data. I personally use many search engines and have to admit that, to this day, Google is still the best one in terms of the quality of answers ... or not. I have just noticed how social media, I am thinking specifically at Twitter, can provide much deeper and insightful results.
The social media Centaur
Social Media (SM) are barely considered as being social semantics search engines. There are SM platforms specifically tuned to this idea: Pinterest, Quora, Reddit etc. The interesting feature of Twitter is that it mixes standard SM aspects (pictures of the dog, of hollidays, inside jokes, chit chat between friends etc.) with this social search engine (SSE) aspect. You often have opened question like : " What do you think is interesting in X?". It is interesting because there is an interplay between social bounding and the effectiveness of the research. People who are going to see your tweets are also the ones with whom you have interacted. They will answer more often if they “know” you and have a shared history of interactions. Lets say it is a virtual variant on helping your eldery neighbour with a large grocery bag. But how do you know those people in the first place? There are “celebrities”, in the last century acception of the word, of course, but most of them have been presented to you via an Artificial intelligence that gathered your searches and interactions and thought that this was a good person for you to be aware of. You have this strange thing that you need a search engine in the first place to find interesting people and to build your feed. It looks ike using a small crane to build a larger one. In this sense the Twitter SSE is a centaur: half-AI and half-human.
Scientific communication in a digital era
The problems of peer review editing are well-known. Nowadays with Nature publishing articles on Britney Spears while we are in the midst of a pandemics, it is just blatant that things have gone awry beyond words. Scientific communication needs to move on from an outdated model. At the origin of this model were the technical difficulties to broadcast information: you need to actually print, and then spread the books/journals/articles. It is very costly. Therefore you need an editor that vets very seriously what is going to be printed. Television and radio partly alleviate the cost (because it doesn't matter how much people are listening, ie the marginal cost is 0) but they lack permanence and require synchronicity. You need to schedule your day around the transmission. All those barriers have fallen.
When moving from an oral to a written culture the skills required to acquire knowledge changed. In the antique there were very sophisticated mnemotechnics to gather as much information as possible. The fact that Iliad was written in verses is a testimony: rhymes help you to remember (and many other testimonies in how the text is written to help memorization can be found). When the culture become written you can outsource your memory. The new skill is to know where to look for: librarians have developed all sorts of ways to do this. In a digital era some searching activities can be outsourced to machines. It works very well for syntactic searches: basically looking for the occurrence of a given word in a text. It is another story when you try to search/crawl ideas. The semantic web never worked as well as T. Berners-Lee thought it would. This is where SSE becomes interesting: it could provide the tools for an efficient way to navigate ideas and avoid drowning in a see of pointless messages. I think there is room for interplay with new crypto technologies because they natively integrate the idea of authorship. Moreover a good incentive structure could be design to help to find out interesting material. Something more subtle than just counting “likes”. You could imagine having a crypto market in which you can bet on ideas/papers. The fact that you cite a piece of work should grow its reputation etc. You could imagine a market of scientific ideas in which peer review is distributed among readers and academic world. The fact that you cannot perform censoreship on one hand, and change what has been published on the other hand would be great. You could have academic freedom together with accountability in the form of reputation and unmodifiable track record (so if you just write nonsense you will gain no traction).
Discussion vs reading
The social media aspects reintroduce the idea of discussion that was lost with books/texts. By nature SM are interactive. So when you are searching ideas in an SM platform you have interactions that are very different from the ones you have with Google search. Your virtual friends may point you to interesting things you did not thought about, or they can react in a way that makes you think deeper than what you would have achieved on your own (maybe they consider the question from another angle etc.).
Clearly we do not master this kind of media yet. Today we are more focused on the problems raised by the echo chambers and the antagonizing effects of SM. Those are not superficial critiques but they are hardly the whole story either.