In my first job I was employed as graduate level clerk providing what could be considered to be a human "chatbot" service to technical management in a research organization (The Cooper Technical Bureau) of an international company (Cooper. MacDougal & Robertson, later part of the Wellcome Foundation). The information department acted as a combined library and mail room for technical correspondence and my job, as a member of a small team, was to ensure that management (both in the UK and overseas) were fully informed of issues which could affect the development of the veterinary and insecticidal products that we sold worldwide. As such I was taking information from manually indexed text documents and providing summary repost and answers to question. I was also involved in integrating new information into the existing paper archives.
Draw an office with the far wall being a large window through which can clearly be seen two or three ostriches in an African savannah landscape. In the office a man is sitting at a desk writing, with a small pile of papers beside him. On the other side of the office there is a four-draw filing cabinet and books in a bookcase.
I ended up with responsibility for correspondence with Africa, North America and Australia and in one way I was in a similar position to a modern computer chatbot in that, for many of the topics covered I had no direct personal knowledge. My office had a window which looked out onto a wide English lawn and sometime my mind was far away as I assessed correspondence concerning the need for better ostrich medication for farms in South Africa, a country which I have never visited, I could not picture (particularly as I have aphantasia) what an African ostrich farm would look like, or what the problems would be when handling diseased ostriches. On other occasions it might have helped me understand the correspondence if the imagined window could show me cattle wallowing in the mud on the shore of Lake Victoria, or of cattle infected with ticks being sprayed on a farm in Australia.
This problem of putting information into context relates to the criticism that all AI packages must be limited because they don't have human bodies or experiences of the real world. I was often dealing with technical correspondence relating to issues where I had no first hand experience to draw on, and it was only by chance that I had some real knowledge of UK farming. The important thing was that the correspondence was between two specialists who knew what they were talking about. From the way the texts were structured I could almost always identify what the key technical terms were and what problems needed to be solved, without needing a lot of background knowledge or a visual image to help. So my lack of appropriate background did not really stop me producing acceptable management reviews, abstracts or index entries,
However there were cases where caution was necessary and where a modern AI chatbot might have difficulties. When trials were carried out with potential new products in South Africa I had to be careful how I represented the texts in the management reports (which would be copied to South Africa). While the raw data relaying to these trials was almost certainly accurate the stated conclusions were not alway statistically sound. In such cases my reposts would be carefully worded. I would quote the claimed successful result with the perfectly correct addition of the relevant part of the data - such as "sample size 5". This would enough to alert readers of the report to be cautious. It is unlikely that a modern AI computer chatbot would spot this kind of problem. and indicate the problem diplomatically.
Manny of the issues raised, such as a query about the insecticidal properties of the mud of Lake Victoria, proved transient and of no great significance. Others unexpected issues could expand to become "black swan events". Shortly before I arrived the publication of the book "Silent Spring" (now recognised as an environmental classic) led to the abandonment of DDT as an insecticide. When I was dealing with Australian correspondence the report that one farmer had complained that our "successful" new product was no longer working triggered major research. This turned out to be the first reported case in the world of cattle ticks developing resistance to an organophosphorus insecticide that had been introduced to replace DDT. Shortly after I left a report of changes in the ozone layer in Antarctica led to the reformation of the propellant in aerosol products. In a sense part of the work of the department involved exception reporting outside the routine technical area.
After three years I had learnt a lot about how text information is communicated in a large organisation, and to how to humans create and use indexes to retrieve the information. I was also aware of the problems a large company had in working in a complex, and sometime totally unpredictable, market place I started to become aware that computers were being used elsewhere. When my suggestion that computer might be relevant to my work was turned down I simply moved to a nearby computer centre, to work on processing the sales of a major oil company - so I could learn how to produce better management information. This decision was, in essence, the trigger of the CODIL research project.
No comments:
Post a Comment