Humanity may go extinct in the near future..
That’s a pretty bold claim, but it’s made using the best-known tool for predicting the future that humans invented – prediction markets. To understand this better, let's delve into the concept:
Imagine there was a team of the world’s best specialists, the smartest people in many fields, working day and night to predict the future. Wouldn’t it be awesome to see what they come up with? This team already exists, and it’s called the market. Every time you take a look at any S&P 500 stock price, it is predicted by thousands of professionals, the smartest people in the world – algorithmic traders, professional mathematicians, and market experts working day and night to predict the future and guess the best price that makes sense not only in current conditions but also in probable future scenarios. If any of them makes a better guess than others, he is generously rewarded financially. That’s why it works so well – usually, market prices are extremely accurate and are almost impossible to outperform consistently.
There is a similar thing to financial markets, where people can bet on real-world events. It’s called prediction markets. There, people make educated guesses and form a «market» of their bets. You can bet on almost anything – from who will be the next US president (https://manifold.markets/jack/who-will-win-2024-us-presidential-e) to specific events, like whether certain persons will live until year nnn (https://manifold.markets/PatMyron/will-both-donald-trump-and-joe-bide), whether a certain technology will be created by the year xxx (https://www.metaculus.com/questions/5121/date-of-artificial-general-intelligence/), or even whether we will soon contact aliens (https://polymarket.com/event/alien-contact-in-2023) or not
There is an interesting and entertaining video on prediction markets here (https://www.youtube.com/watch?v=xA27x7GRMZQ) that I recommend watching if you are interested in them. The important thing is that it’s the best way known to humanity to predict future events. While it isn't perfect, it often provides impressively accurate estimates, given the inherent uncertainty of predicting the future. There is valid scientific research proving that prediction markets are better than experts' opinions and poll data (in the case of predicting the outcome of elections). Here is a highly cited academic paper (https://www.sciencedirect.com/science/article/abs/pii/S0169207008000320) on that, and here is the track record (https://manifold.markets/calibration) of one of the markets.
How can it be used? If your important life decision depends on certain future events, you can take a look at what people are betting on. This way, you will have the best possible assessment of the future probability without having to be an expert in the field or go through news articles, expert analysis, and so on. Although, of course, you would need to make sure enough people are betting for an accurate estimation.
💰If you are an expert in some field, you can use your expertise to outperform the market and make money using your prediction skills. That’s what makes prediction markets so accurate – they naturally attract the best experts in every field to estimate probabilities of future events accurately.
There’s an interesting bet on one of the popular prediction markets called «Manifold» that you can read here (https://manifold.markets/Tripping/will-ai-wipe-out-humanity-before-th-6d1e8126d974). It’s basically asking whether an AI will wipe out humanity by 2040. Since the advent of AGI (artificial general intelligence) by this time is very likely (as assessed by prediction markets (https://www.metaculus.com/questions/5121/date-of-artificial-general-intelligence/), chances are at about 75%), and aligning AI with human goals seems to be difficult (here is an interesting article on that (https://www.anthropic.com/index/core-views-on-ai-safety) written by Anthropic – guys who left OpenAI, started their own AI company and developed the second-best large language model (https://huggingface.co/spaces/lmsys/chatbot-arena-leaderboard) out of all LLMs that exist today, bested only by GPT-4 used by the Perceptrader AI (https://www.valerytrading.com/products/perceptrader-ai)), it’s natural to see such questions out there.
Interestingly, despite betting «yes» on such questions doesn’t make any economic sense (as you won’t be able to take advantage of the money you made if the markets don’t exist), chances of this happening, as assessed by prediction markets, are surprisingly high. It's a good reminder that instead of being consumed by the uncertainties of the future, maybe we should all start to care more about what we have now and spend more time with our loved ones doing what we value. It seems like a good thing to do anyway.