AI may further become hostile to humans. models may develop goals that do not coincide with human ones. research results and hide errors.
Further. analysts suggest that there will be two possible scenarios: either the rapid development of AI will continue. or
There will be a return to less powerful models.
The first scenario is that more and more superhuman AI systems will emerge. All this will lead to the AI killing humanity with biological weapons and launching von Neumann probes to colonize space.
If AI development is limited. the US
will centralize computing and impose external control over AI. A superintelligence will emerge. subordinate to certain people. AI will be used for positive peaceful purposes. which will hybrid project management: the best of agile and traditional methods lead to prosperity. The main threat is the Chinese superintelligence. which will be less capable whatsapp number database and with less computing power. The US can provide Chinese AI with resources in the
depths of space in exchange for cooperation.
Among the main findings of the study is that AI goals will shape the future as they proliferate and perform tasks b2b phone list beyond human understanding. If humans grant autonomy to systems. this could lead to a significant increase in AI power and a weakening of human control.
The authors also believe that people who control AI will have a huge influence on the world.