OpenAI is warning about the dangers of runaway AI systems, even while it competes with other major tech developers to build "superintelligence" -- an as-yet theoretical machine intelligence that outperforms the capabilities of the human brain. In a blog post titled "AI Progress and Recommendations," published Thursday, the company outlined its vision for the broad-scale social benefit that such an advanced AI could confer upon humanity, the risks that could be encountered along the way, and some suggestions for mitigating them.
Reasonable people from all sides of the current AI conversation agree that this particular paradigm shift is just different. If you don't believe that, record 20 minutes of your own stream of consciousness with an application like Otter and ask it to summarize your thoughts. Or download RunwayML and have it turn photos of your friends into videos of them doing the tango.
After decades of research, there is still no clearly articulated scientific consensus on what sleep is or why it exists. Yet whenever sleep comes up as a topic of discussion, it is quickly reduced to its necessity and importance. Popular media remind us of what can, and will, go wrong if we do not sleep enough, and serve up some handy tips on how to overcome insomnia.
Suleyman's "central worry" is that SCAI could appear to be empathetic and act with greater autonomy, which would lead users of SCAI to "start to believe in the illusion of AIs as conscious entities" to the point that they advocate for AI rights and even AI citizenship. This would mark a "dangerous turn" for society, where people become attached to AI and disconnected from reality.
Since 1990, there has been a sharp decline in how many people men say they are close to, and two-thirds of men aged 18 to 23 think nobody really knows them.