Hello Andrew,
I was reading the web page on "A Gentle Introduction to ML". It seems to me that whoever answered the questions in "Some objections to ML" has had very little experience in the industry. Some of the claims are totally unjustified. There is no good reason to believe that ML is the panacea. Where is the evidence to show that ML is the best possible language to be adopted by all. Many studies (in formal methods) have demonstrated that a particular choice of a language contributes very little to the correctness of a program.
I am sorry to note that the justifications made seem to be more an emotional outburst rather than an objective comparison. What would be more attractive and tasteful is to put up an objective assessment of why somebody should follow a particular approach.
The use of distasteful language such as "cartels" or "Industry flirts" will only be counter-productive.
Finally, let's not forget that technological marvels such as spacecrafts or even super fast trains have been realized in industries. They couldn't have achieved this had they been "arbitrary" as your article suggests.
cheers
Manju
I do have a little experience of industry. I worked as a contractor for several years before being elevated to this ivory tower. I have many students who work in industry and study part time.
I do not claim to represent the industry nor academia and I deny anyone else who does. Moreover I concede that I have been accused of being out of touch with the "real world" both as a lecturer and as a worker in the industry.
I thought the "industry flirts" metaphor very tasteful. I had much earthier language in the first draft.