No advantages accrue to the chemist from knowing how to generate and how to interpret a chemical code. Codes are needed only for the mechanical manipulation of chemical structures. Clearly then, if the coding of chemical compounds could be accomplished automatically, this automatic conversion would relieve the chemist of a considerable burden.
The success of any new molecular encoding method relies, in part, on its invisibility to its prospective users. After all, why should anyone bother to learn yet another molecular language, especially one designed with computers in mind? Yet these encoding systems are critical in connecting chemical information and information technologies. How can any new encoding method be made part of existing workflows invisibly?
Feldman and his group at Walter Reed faced a similar problem in the early 1960's. American Cyanamide had been using a modified typewriter to prepare attractive 2-D chemical structures, purely for human consumption. Feldman's idea was to modify the typewriter design still further such that a computer-usable molecular code would be recorded as a byproduct of preparing the structure diagram. The typist could remain blissfully unaware of the mechanical magic beneath, and get on with his or her job. The idea was later adapted by Shell to produce a more cost-effective device.
The structure editor has long since replaced the chemical typewriter. But the same forces are at work with today's new molecular encoding methods, especially InChI. To what extent are scientists themselves being given the tools to leverage these new technologies, without having to become aware of them? What will these new tools look like and how will they differ from what came before?