Last week, ChatGPT revealed that it was better at being me than I was at being me. Fortunately, so far, it is only part of me where artificial intelligence is better. I had intended to write this weekly column to update you about the need to abolish the capital gains tax.
A colleague, who is a fine economist, had been doing work on monetary issues and mentioned to me that by using ChatGPT he was able to improve and speed up his work product. So I decided to give it a try, and I asked ChatGPT to “write a 700-word essay about economist Richard Rahn’s views on taxing capital gains.” The result was astounding.
Within seconds, the computer wrote a very fine essay, correctly presenting my arguments and those of critics, all drawn from the many papers and articles I have written on the topic over more than four decades. I then asked it to write an essay about my work regarding the size of government and economic growth. Again, within seconds, I was presented with an excellent essay — perhaps better and more complete than I could write now — even though it was all based on things I had written over the years.
After experimenting further with ChatGPT, I concluded that it is useful and reasonably accurate in dealing with questions and topics where there are many papers for it to draw from (such as the problem with capital gains taxes or the optimum size of government), and almost useless and not very accurate in dealing with topics that have limited source material, such as t.
