This should get a runner-up in the Darwin awards.

Some time ago I saw a very interesting lecture by Douglas Hofstadter to the effect that Ray Kurzweil is a crackpot, and his whole Singularity notion is a crackpot one. However, in language of great respect, Hofstadter warned us not to dismiss his arguments to lightly, but that we need to very carefully examine them to see why they are crackpot. Ultimately he's taken some very sound and reasonable ideas way way too far. I don't mean to say that he's taken them too far because we don't want to believe them, but rather because he is omitting or ignoring other factors. Specifically he's ignoring that:

1) Computation is not exactly the same as intelligence (An iMac can crunch way more numbers than all the neurons in a cockroach, yet it couldn't navigate the walls of an apartment building to find food or mates), which is not exactly the same as consciousness (Your tivo may be able to predict that you like Battlestar Galactica - but can it understand why you like it?), and:

2) that just because we've seen an exponential growth in computational power up until now doesn't mean that trend will persist (rabbits may show exponential growth, but there are other factors which may arise to limit it - ie lack of food, rise of preditor populations).

Borromean rings? Aren't these the ones that Boromir smashes apart with his sword? Maybe I'm getting confused.

WRT RyanH, I agree that on principal a straight-up replacement of existing web technologies with 3D-world versions adds only gimmicky "neato" value, and nothing more sustained. To expand a little bit. I'm not a UI expert, but imho a good interface is about creating the right metaphor for the medium. A good interface should enable you to use the full power of a medium as easily as possible. With communication tech, the ideal is that a user should be able to do something as quickly as the desire is formed. The fewest steps to doing it in as unambiguous a way as possible is the best approximation to that ideal.

A 3D world inherently adds barriers to getting most communications done - because to sustain the VR metaphor you add constraints to action (you have to move to a particular location, and perform the right physical actions). This is desirable in a game like situation because it is the interaction with the world which is itself the desired objective. As a UI onto some communications app, it is inherently bad, because it makes that layer less easy to use (more steps to accomplish the same goal). Furthermore, by allowing every user to change the UI metaphor in his/her own world, you make it even harder in that a visitor must learn the particular semantics of every new world they browse.

However, this doesn't mean that a tight VR-web integration is a bad thing. What any easy-to-use VR dev. kit type environment (whether it's game modding, 2nd life or metaplace) should aim to be is a platform for new types of interaction - not a replacement or front-end for existing ones. Many commentators poo-pooed the web in the mid '90s as gimmicky and pointless - I don't think they ever foresaw googlemaps, wikipedia and boing-boing. We should be cautious of dismissing platforms like this just because they are as-yet immature. I believe there are whole new ranges of social interaction being organically developed online right now in VR worlds (who would have foreseen the rising popularity of machinema?). The thing we need to look for is some way of formalizing these new modes of interaction into something like a design philosophy and embodying that in something like a toolkit. Who knows - ten years from now we may all be building sites with the VR equivalent of AJAX!