I recently published an article about using AI to mark students’ work. As I wrote there:

I didn’t think that answer was good enough. I didn’t ask how ethical the proposals were. I asked it to mark an Economics essay. In retrospect, I should probably have provided a rubric of sorts (even though I don’t like rubrics), in order to focre it to mark the essay in accordance with Economics principles.

However, all I wanted to do for this follow-up was point out a couple of things that are quite important in this context.

Firstly, in England we have what Lord Hailsham once described as an elective dictatorship. In theory, at least, a government can do what it likes, especially when the governing party has a huge majority, as is the case now.

Secondly, literally a couple of days after I published the article there was a piece in the news about how the Chancellor of the Exchequer was being urged to impose a huge exit tax on millionaires wishing to relocate abroad. The background to this is that we are being promised huge tax increases in the areas, especially, of capital gains tax and inheritance tax. Unsurprisingly, at the last count 9,000 millionaires were in the process of leaving the country.

I’m not interested in discussing the ethical, economic or political aspects of this here, but just to point out that for Perplexity to dismiss my proposals as “problematic and unethical” was very disappointing because it simply sidestepped the issue. That issue was very simple: from an economics standpoint, could these draconian proposals work? The answer is that yes, from a purely theoretical viewpoint, it’s conceivable that they would. What the AI application “thinks” of them from an ethical point of view only serves to unveil its own built-in prejudices.