As We May Think -- Vannevar Bush

Article here.

Bush's prophetic 1945 article outlines a postwar environment where information is plentiful but methods for collecting, storing, indexing, and retrieving are inadequate. He proposes that there must be a shift in scientific efforts from controlling human physical environment to making human knowledge more accessible. Where's an internet when you need one?

Bush discusses several tools that exist and ones that *should* exist, most of them focused on data recording (image and sound) and storage. He notes that scientists are not the only people who manipulate data and examine the world around them and that all people can profit from the "inheritance of acquired knowledge." But if only a limited amount of information is available, people cannot use it to their full potential. However, because of the way people naturally organize information by association (instead of numerically and alphabetically), new methods must be considered. Bush discusses a "memex" -- a computer system implemented with electromechanical controls and microfilm equipment, that would permit a researcher to follow and annotate topics of interest, analogous to later hypertext technologies (from wikipedia).

Some interesting points:

• The parallels Bush draws between the technology and human anatomy. The technology would be most effective if it can mirror the way branis operate. He poses the technology as an "extension of the human brain" that follows "trails" much like the brain does but with much more clarity and permanence. What's even more interesting is that this type of indexing has yet to be fully explored -- we still interact with technology, especially computers, largely in unnatural and illogical ways.
  
  
• Speech recognition -- "talk to type": what has kept this from catching on? I play with it on my computer all the time, but I never think to "write" a paper with it. Too inaccurate?
  
  
• How far long are we...really? I used to work for a content management software company that enabled users to input, index, store, retrieve, share, workflow, and destroy information. Sounds pretty cool. Who wouldn't use that?
  
  The product was solid. We won many awards and had a partnership with Microsoft. Yay, big money. Except no. Because at the end of the day, content management software is only as good as the users inputting the information, regardless of how easy the input steps are.
  
  Sure, OCR could help users auto-index items and fields could be set so information us automatically pulled from certain types of reports and documents. But if Joe in accounting doesn't tag the report as a mortgage note or Jane in finance doesn't approve the P&L sheet in the workflow and send it to the auditing team, the software is useless.
  
  In this regard, Bush's proposal for technology to closely mirror the way our brains are wired make a lot of sense. When we interact naturally with tools -- the big Google search box comes to mind -- we're more inclined to use them more often. But storing and indexing are huge undertakings, regardless of task. Just consider how many times you've created a document that required collaboration. Think about your naming conventions and how you organized the content. Now think about how others try to understand it and make it their own, and you end up with multiple versions of the document. Multiply that by hundreds for some industries -- hundreds of thousands for, say, research and financial industries.
  
  Who will organize, index, and generally make sense of this? Not the technology.