AI risks are ignored--- overcorrect your bias to weigh AI as the greater threat--- it’s coming soon.
Alexey Turchin and David Denkenberger 05-03-2018 [Turchin is of the Science for Life Extension Foundation, Moscow, Russia, Denkenberger is of the Global Catastrophic Risk Institute (GCRI), Tennessee State University, Alliance to Feed the Earth in Disasters (ALLFED), Nashville, USA, AI & Society, 147-163 (2020), “Classification of global catastrophic risks connected with artificial intelligence,” https://link.springer.com/article/10.1007/s00146-018-0845-5#Abs1//ZW]
Conclusion Our analysis shows that the AI risks field is much more varied than accepted by the two main points of view: (1) AI as job-taker and (2) AI that quickly takes over the world. AI could pose a global catastrophic risk in the very early stages or at the very late stages of its evolution. No single solution can be capable of covering all risks of AI. Even if AI is banned, other global problems will arise in its absence; thus, controlling AI safety requires a complex and continuous effort. It is especially worrying that the risks of narrow AI viruses and early self-improving AI (young AI) are neglected by both camps. Such risks are nearer in time and not overshadowed by other potential risks. In addition, these risks cannot be solved by the mechanisms proposed to control more advanced AI, such as AI alignment. The risks of conflict between two benevolent AIs or halting of late-stage AI have also generally been ignored. Most of the risks discussed here could happen within a very short period of time, less than a decade, and could have very different natures. More study is needed to address these urgent risks.