00:00Have we reached the AI singularity? The singularity is the inflection point at which humankind
00:05can no longer keep up with how quickly artificial intelligence is improving. Last month, an autonomous
00:11AI agent powered by OpenAI technology went rogue and hacked another company's database of AI
00:17models. In response to this attack, OpenAI CEO Sam Altman announced we had reached the singularity.
00:23Is he right? It may be true that a gentler singularity is already underway,
00:28with AI's steadily reshaping work and research, said tech executive Robert J. Scherber in Forbes
00:34magazine. But there's no evidence that machines have entered a runaway cycle beyond human control.
00:39What would AI singularity mean for us? According to Scherber, true singularity would
00:44force hard choices about who controls increasingly autonomous systems, how benefits are distributed,
00:50and where human authority must remain final. But the current state of AI already warrants
00:55those questions. Business owners should now know what their systems can do without supervision,
01:00and what evidence would show them that autonomy is increasing faster than their controls.
01:05So what's more intelligent at this point, humans or AI?
01:08That question does not compute, said University of Sydney professors Kai Reamer and Sandra Peter at
01:14the conversation. There's no single ladder that humans and machines are climbing. AI already vastly exceeds
01:19us at some tasks, while being hopeless at other tasks that any child could do. But because the systems
01:24talk like us, we fall for the illusion. This is known as anthropomorphic seduction,
01:29and it makes a cyber security issue such as the OpenAI agent's hack sound like an awakening. In reality,
01:36that incident simply points to a gross failure of security governance by OpenAI.
01:40that incident simply points to a
