The various teams were compared to the prediction market, and in all three cases-- humans, AI, and human-AI teams-- the prediction market was better*. So... if the predictions in the prediction market are better...what other possible prediction method did they use that wasn't humans, AI, or human-AI teams?
There are two reasonable answers to this. One is that the prediction market includes some fraction of investors that have insider information, and hence the prediction market does better because some of the investors know the answer before investing. The second is that the people in the investment market are professionals at this, and self-select to invest in things they have studied and already know something about, while the teams recruited for the experiment were amateurs who didn't know anything to start with.
I certainly hope the second explanation is the right one.
--
*(yes, the article says 8-10% of the teams "rivaled" the prediction market's accuracy. But that tells us that 90-92% of the teams failed to rival the prediction market's accuracy. So, no, human AI teams didn't beat the market. And if you post-facto examine the results and pull out from a group a subgroup that does better, that's p-hacking, not statistics.