3 ms·
I think that the hard part of tackling these ambitious projects is (often) not the actual engineering but rather making the ideas popular and fighting against t
by ilaksh 15y ago
I think that the hard part of tackling these ambitious projects is (often) not the actual engineering but rather making the ideas popular and fighting against the status quo. Another hard part is how difficult is the transition from how we do things now to the new way.
So most of these things he mentions, people are working on them, or something similar, or even have functional software. That software just isn't popular. Not because it isn't useful, but because it didn't catch on.
And a lot of these ideas aren't really useful until they reach a critical mass of users, which makes it even harder.
The big ambitious thing I wanted to mention was DONA (data-oriented networking or http://en.wikipedia.org/wiki/Content-centric_networking http://en.wikipedia.org/wiki/Content-centric_networking). You might be able to combine that with some type of semantic knowledge storage and engineering along with a type of e-democracy. People have working examples of these things, its just hard for people to pick them up and start actually using them and then mention them to others for them to become trends.
Before (or instead of) human-controlled knowledge engineering we may see Google (http://mashable.com/2012/02/13/google-knowledge-graph-change-search/ http://mashable.com/2012/02/13/google-knowledge-graph-change...) (or possibly some start up) come out with a Watsonish system that builds huge knowledge graphs by actually comprehending the semantics of web pages it spiders and then lets you query them more naturally. (Which I guess that type of system does exist, just didn't catch on, maybe because it wasn't quite up to human level comprehension or didn't become popular for whatever reason.)
- disgruntledphd2 15y agoThe KDD cup had some entries in 2006 on this idea. I found one of the papers by accident yesterday. It looks quite interesting, might be worth checking out. www.daimi.au.dk/~ifrim/publications/kdd2006.pdf
- waterlesscloud 15y agoOne of the benefits of something like Y Combinator is that if you have one of these ideas, and you can actually make it work, YC and its associated network of people can be massively helpful in spreading the word. In fact, that may be the major benefit of YC.