From the CIA Teams discussion at work after I shared a link sent to me by Paul K. (slightly edited):

Joe: I have not read all of it yet, but there are some “interesting” projections on AI here: AI 2027.

I want an underground bunker in Idaho.

Sean: (my manager): Got room for more?

Joe: Sorry, all full and then some. I have three children, two step children, their spouses, and five grandchildren.

Sean: What if I can provide valuable skills?

Joe: Two son-laws are army combat vets. One is a world class (#5 or #6) shooter. Stepdaughter is a world class machine learning expert. Stepson and one daughter are top notch computer programmers. The IQ range of the kids and step kids ranges from 125 to at least 155. My brothers are farmers two miles away with non-computerized farming equipment, fully equipped metal and wood shops. One daughter has an accounting degree. There are two art majors in the mix.

We have more guns than we can count. Probably 50,000 rounds of ammunition. Equipment to assemble two to three thousand rounds per hour (given the components).

What are you bringing? We could use someone with high level medical training.

At this point if you haven’t been planning and working on it for several years and the predictions come true you probably are too late.

I’m not convinced the predictions are likely to come true. Historically, there have been many, many predictions of doom. While there have been lots of disruptions of “normal” life, things always resolved into something pretty good. The examples include industrial revolution, nuclear weapon proliferation, nuclear winter, global cooling and more recently global warming. Women’s liberation and the birth control pill were used to predict social catastrophe. Before that, women being allows to vote was predicted to be catastrophic. Even the Internet and video games were predicted to be disruptors of society. 

Read the books from the 1960’s and 1970’s such as Silent Spring (I was convinced when I read it but later found out that many of the references were fake and/or severely distorted) and Future Shock (required reading for an engineering class in college, I did not find it believable and was very annoyed with the movie version). They were very popular in their time but dismal failures at predicting the future.

Fear sells and spreads far better than predicting increasing or even maintaining our current comfort level. Looking at the probabilities of propagation, you are far more likely to see doom predictions than the utopia predictions. But the long-term historical trend lines by far favor the utopian vision.

That said, the way you want to bet is not necessary on what is most likely to occur. If you cannot afford to lose the bet but betting in the direction you likely have a moderate outcome in nearly all situations you should give serious consideration to betting on the lower odds outcome.