EROS, Artificial Intelligence and Expert Systems

In view of the current interest in Al and ES across many disciplines as well as in chemistry, it seems appropriate to give our views on how EROS fits into the scheme of things. 

EROS is not able to do this. In fact, it falls down in its inability to extract its own rules from input organic reactivity data sets. It is this information to which organic chemists have applied their intelligence and experience, and then, in the case of our research group, built into EROS. Thus, in the sense which we believe was intended when the term artificial intelligence was coined, EROS is not yet an example of Al. 

We are therefore inclined towards the opinion that EROS can be classified as an ES, based as it is on distilled knowledge of organic chemistry in the form of our quantitative models (Sect. 8). EROS calculates its results based on these models, in contrast to the organic chemist who depends on a sort of mental qualitative pattern recognition when considering functional groups and their interrelationships. Despite this difference, EROS can provide answers to questions that are not always straightforward even for an expert in the field. 

We have dwelt on the semantics of Al and ES merely because of the current interest in such matters. However, we conclude this article by emphasizing the fact that EROS functions successfully as a tool to aid chemists, wherever it fits into the taxonomy of computer software. We believe that this is far more important than whether the system conforms, or not, to any particular definition of Al or ES. 

---