Just in case you wanted a serious answer to your question... in the engineering profession we spent a whole lot of time looking for information. Including: * Finding books/manuals with the right formulas, then looking up the formulas in the book * Finding data, eg. "latent heat of evaporation of methanol" * Finding installation/operation/maintenance manuals. We had a massive library full of manuals. In many cases, the right resource was not on hand, and you'd have to find the person who had it or knew where to find it. Many times this had to be done via correspondence (fax by then, thank god, not postal mail) Worst of all, many times you just had to generate the data yourself, by experimentation. This included things like actually measuring the latent heat of evaporation, for example, or trial-and-error wrt eg. disassembly or maintenance procedures. Often "stuff" was designed far more conservatively, to compensate for lack of data. The internet has been a stupendous boon to engineering productivity, and largely generated the free time that some people spend now on Facebook etc.

If you don't get your work done, you lose your job, so whether you have internet access or not doesn't much matter. You will regulate your own distraction or you will perish. That being said, there can sometimes be lots of downtime in office work. I imagine that the internet is there because it LOOKS a lot like doing actual work. It's much easier for the employee to pretend he's working, and it's much easier for the boss to pretend she hasn't noticed her employees are being unproductive. A boss that sees her employee spends four hours a day chatting at the water cooler has to "address the problem" to save face. If her employees stay in their cubes she can lie more convincingly to her superiors that everyone in her department is quite busy. It is simply covert idleness. Just channeling my inner Hanson. (for what it's worth, I do all my LessWrong reading/posting from work)