Results from The Truth Test
📆 September 27th - October 27th
🌐 Crowd size: 21 participants
Goal: test a new approach to crowd wisdom that can outperform experts at predicting things about the real world.
We predicted the 2024 Election outcome on October 27th using a crowd size of 21 diverse participants.
Neighborhood safety was a big topic during the 2024 election cycle.
We discovered that the downward trend in safety was largely overstated by the crowd and likely isn’t as significant as people think.
Social media’s popularity can obscure how much it actually helps people with personal relationships.
We found that only a small fraction of people get relationship value from socials.
Surprisingly, almost no one is satisfied that they have the best access to information.
The vast majority of people do not believe experts are enough; and they doubt both the intent & accuracy of new information online.
There is a bias to believe others feel satisfied with new data online and that they are getting exclusive access, when in fact most people don’t feel this way.
The crowd predicts AI will not take their jobs for at least 10 years.
This confidence is surprisingly popular.
Let’s let the crowd rank influential opinions:
Accurate and reliable according to the crowd:
TruTake’s result is more in line with a perspective from Daron Acemoglu, a professor of economics at M.I.T., who said on October 3rd, “I don’t think that the most optimistic views in the industry, which people sometimes refer to as ‘scaling laws’ actually will hold water when you look at it over the next 5 or 10 years.”
Middle of the ground:
One of the 2024 Nobel Prize in Physics winners, Geoffrey Hinton, said on October 26th: “Until very recently I thought [the existential threat to the human species caused by AI] was a long way off. I use to think it would take 50-100 years before we had really smart things. Now I believe within the next 20 years we will have something smarter than we are.”
Not accurate according to the crowd:
For contrast, consider Elon Musk’s latest comments at Riyadh: “I think [AI] will be able to do anything that any human can do, possibly within the next year or two… The question is how long before it can do what all humans can do collectively, probably 3 years from that point.”