I came across this interesting debate while reading a paper by Andy Oram:

Two ideas, diametrically opposed in philosophy and approach, have seized the attention of Internet companies and technologists over the first few years of this century. (...) But these two ideas are attracting both money and attention. One stresses classification, the other community. Neither has borne much fruit yet. (...) The first idea goes by such names as the "semantic web"—coined by Tim Berners-Lee, the creator of the World Wide Web—and "Web Services." It leads to infinite meetings of standards bodies, taking up hours of valuable technologists’ time, who report year after year on the tremendous progress they are making toward ever receding and ever more audacious goals.

The other idea goes by the name "social networking," and brings out breathless talk of a revolution in social relationships, supposedly to be opened up by "frictionless connectivity": the ability to find almost instantly the person who can meet your specific personal or career-related needs. After several years of experiments in each area, outlines are emerging for the domains where each idea may prove valuable.

The article also expresses critics about social software: "The public has mostly lost interest with the first wave of sites that offer social networking, probably because what they offer seems to add little except extra overhead to current Internet services such as email and newsgroups.". I think it's definitely true, my ultimate social software is my blogroll. I like this kind of statement: "I expect most members of online social networks are as inactive as I am, having tried them out and been unimpressed." because it definitely summarizes what I felt after trying those tools.

But the author is still positive on it:

These criticism apply to social networks they way they’re currently implemented. Because viral marketing and new media have an excellent possibility of becoming important social movements, I think online social networks will grow in importance, and at some point somebody will make one that works. We can also move to yet another stage where we statistically measure our network and learn from aggregate facts about the people we know. There’s plenty there for a century of innovation