> 1. People who have insider information, or at least more sophisticated predictive capability than your average person.

This bucket as you've defined it is too broad.

There are a few different kinds of non-gambler participants in prediction markets:

1. People with "insider information" as we think of it - they "know" the answer to the market because they are "involved" somehow.

2. People who aim to do superior analysis of publicly available information to produce an edge. For example, an AI firm with better hurricane prediction modeling may try to monetize that by betting on whether or not a hurricane will impact an area.

2b. People who do the work to create new information. For example, the Trump 2024 election market on polymarket famously had better odds for Trump than polling. It turned out that a mega whale was bidding Trump up because he had paid for his own private polling in battleground states and that gave him confidence Trump was going to win.

In short, it's mostly incorrect to suggest that prediction market participants are either illegitimate insiders or gamblers; there is a third class of actors that are a very important cohort: those who do the work to create better predictions and monetize their work by betting in the markets. This third cohort of professional predictors is the most important in long-term prediction market growth.