Editing Memex (section)

==1967 Memex revisited==
In 1967, [[Vannevar Bush]] published a retrospective article entitled "Memex Revisited"{{sfn|Bush|1967}} in his book ''Science Is Not Enough''. Published 22 years after his initial conception of the Memex, Bush details the various technological advancements that have made his vision a possibility. Specifically, Bush cites photocells, transistors, cathode ray tubes, magnetic and videotape, "high-speed electric circuits", and "miniaturization of solid-state devices" such as the TV and radio. The article claims that [[magnetic tape]] would be central to the creation of a modern Memex device. The erasable quality of the tape is of special significance, as this would allow for modification of information stored in the proposed Memex.{{sfn|Bush|1967}}

In the article, Bush stresses the continued importance of supplementing "how creative men think" and relates that the systems for indexing data are still insufficient and rely too much on linear pathways rather than the association-based system of the human brain. Bush writes that a machine with the "speed and flexibility" of the brain is not attainable, but improvements could be made in regard to the capacity to obtain informational "permanence and clarity".{{sfn|Bush|1967}}

Bush also relates that, unlike digital technology, Memex would be of no significant aid to business or profitable ventures, and as a consequence, its development would occur only long after the mechanization of libraries and the introduction of what he describes as the specialized "group machine", which would be useful for the sharing of ideas in fields such as medicine. Furthermore, although Bush discusses the compressional ability and rapidity so key to modern machines, he relates that speed will not be an integral part of Memex, stating that a tenth of a second would be an acceptable interval for its data retrieval, rather than the billionths of a second that modern computers are capable of. "For Memex," he writes, "the problem is not swift access, but selective access". Bush states that although the code-reading and potential linking capabilities of the rapid selector would be key to the creation of Memex, there is still an issue of enabling "moderately rapid access to really large memory storage". There is an issue concerning selection, Bush conveys, and despite the fact that improvements have been made in the speed of digital selection, according to Bush, "selection, in the broad sense, is still a stone adze in the hands of the cabinetmaker". Bush goes on to discuss the record-making process and how Memex could incorporate systems of voice-control and user-propagated learning.{{sfn|Bush|1967}} He proposes a machine that could respond to "simple remarks" as well as build trails{{sfn|Bush|1967}} based on its user's "habits of association," as Belinda Barnet described them in "The Technical Evolution of Vannevar Bush's Memex." Barnet also makes the distinction between the idea of a constructive Memex and the "permanent trails" described in ''As We May Think,'' and attributes Bush's machine learning concepts to [[Claude Shannon]]'s mechanical mouse and work with "feedback and [[machine learning]]".<ref>{{Cite journal |last=Barnet |first=Belinda |date=2008-01-01 |title=The Technical Evolution of Vannevar Bush's Memex |url=http://www.digitalhumanities.org/dhq/vol/2/1/000015/000015.html |journal=Digital Humanities |volume=2 |issue=1}}</ref>