oh god please no. die, web 2.0, die.
It can be implemented in a couple of ways, that aren't limited to the following - generated by most popular topic, most popular users, least popular topic and least popular users. Something along those lines that offers up avenues of discovery. Basically, I'm all for more routes to find subjects here on hubski, as it appears the only way to do that is by adding followers and reading their content, without performing a hashtag search - which can be complicated if you don't know the right subject terms relevant to your interests. And if you really want web 2.0 to die, then you might as well kill off search, links, authoring, tags, extensions, and signals. :)
the tag cloud itself needs to die. It's overdone, unimaginative, of limited use, and terrible UX
The tag cloud is dead only if it remains static. Image a tag cloud of, say, 25 subjects. Clicking on that tag opens up a close up of that tag with related tags. Do this until you've reached the end of the tag. This helps to create a dynamic tag cloud that could be updated simply by added related tags to the original tag. You have to remember that when designing anything you need to try to incorporate elements that appeal to a wide swath of users. Many users will perform keyword searches, some will look up by username, some will want to explore using other means. Tag clouds, in and of themselves, are just another tool of discovery. As someone who has had ample experience working in archives, libraries and with patrons I've come to discover that multiple entry points, that are well designed, coherent and relevant, offer up a variety of means to find the needle in the haystack. All that being said, I'd not like for a tag cloud to overwhelm the design of the site as it is. I like the fresh, barebones look to hubski. I'd just like more ways to discover content. Maybe I'm doing something wrong, but it feels like I'm barely scratching the surface of content here.