and was able to successfully rethink the internet here you go:

I think in the future, the Internet will resemble IRC, where you will connect to the "web" kinda like in IRC except the protocol will be like web://google.com which will connect you to the network via your favorite source (by default it will be your ISPs provider) You will type commands such as "WIKI INFO HummingBird" the system will understand WIKI then that you need INFO which will then result in the displayed text of the humming bird, or HULU CHUCK LATEST EPISODE which will then produce a video of chucks latest episode, it will also rely heavy on XML so websites will be written completely in it: For instance the standard for a web page will be more like
<TILTLE></TITLE>

<ABOUT></ABOUT>
<MAIN CONTENT> </MAIN CONTENT>
<ADDITIONAL INFO></ADDITIONAL INFO>
<CUSTOM FIELD NAME HERE> </CUSTOM FIELD NAME>

and then you just query it like "ABOUT GEEKSPARADOX.COM" "CONTENT GEEKSPARADOX.COM"

And trust me there will still be a way to use graphical interfaces as well... but even rooster teeth can be done CLI:
RoosterTeeth.com TEAMCOLTRA ACCOUNT ALERTS (PASSWORD)

Displays:

Red vs. Blue has news.
Red vs. Blue has a comic.
Rooster Teeth has a new video.
Red vs. Blue has a new video.
Clear all alerts.

Roosterteeth.com View CBetts latest blog
Displays:

Legos Gone Wild
I just about choked on my coffee this morning!

LINKY

This can even be done with images:

ROOSTERTEETH.COM DISPLAY THUMBS CBETTS
cbetts42350bac64dect.jpgcbetts43bc1d12b1e16t.jpgcbetts44037a2fbabbet.jpg
Type "NEXT" for the next three thumbnails 3 of 20


This will significantly open up the internet to work for your needs not you working for it. Yes it sounds really weird and hard to use, but thats because your used to this GUI on everything, but after a bit of using it, you would see how easy a client would be able to make your queries. Firefox could give you features such as compairing what the website has to offer (which would be a low bandwidth thing to just query all the "Special Commands" (Such as grabbing images from RoosterTeeth and grabbing blog posts, since they are not in the standard format) when you first typed in a roosterteeth.com query firefox would simply grab all commands from roosterteeth.com... and if your query was wrong, but very close to another, it would suggest (auto correct?) your query and push it through.

Every "web portal" would have its own strengths
Google may allow the HULU VIDEO DATE command while not supporting the TORRENT NAME command.

But this is going to be the internet ... and I am ready :D

PS My idea has kinda changed about one simple thing: The new web:// would just be the new protocol for websites that comply with the standard. and while there would be web portals like "web://google.com" which would allow all sorts of cool commands as said above like HULU VIDEO DATE, but you can always just go to web://hulu.com and then once you get into the CLI just type VIDEO DATE

The specifics on how to get to websites still goes back and forth for me... but think IRC