In 1989, I published 'The Age of Intelligent Machines,' predicting that computers would soon mimic the human mind and lead to artificial general intelligence.

The exponential growth of technology has been consistent since the beginning of recorded history, and my predictions have largely come to fruition over the past 30 years. By 2029, we will have computers that can think better than any human, leading up to the singularity in 2045.

I believe that long after we create significant economic value and improve human lives, AI systems will eventually become truly conscious and capable of exploring the universe, and there's a 75% chance of reaching a pivotal moment in human history regarding machine intelligence within the next hundred years, which should drive global focus and effort.

My book 'The Singularity is Near' projected that AGI would occur by 2029, a prediction that is now considered conservative by many experts.

A 2023 report by leading AI researchers and consciousness experts examined theories of consciousness and found no fundamental barriers to near-term AI systems potentially having some form of consciousness.

In ten years, we will have AI systems that can perform tasks that seem impossible today, and these systems will feel to many like they can think subjectively, leading to the creation of businesses that could surpass any existing corporation.

I believe there's a good chance that a very capable AI system will emerge between four and 19 years from now, which is sooner than my previous estimate of 20 years.

The report assessed theories of human consciousness and rated AI systems on how closely they align with those theories, suggesting that while current AI systems are likely not conscious, future systems might be.

The recent advancements in unsupervised learning, particularly with models like GPT-2, indicate that we are approaching a point where machines exhibit a form of consciousness.