When it comes to artificial intelligence, no one has more experience than Ray Kurzweil. During his 61 years in the development of AI, he invented the first commercially available large-vocabulary speech recognition software, received the 1999 National Medal of Technology and Innovation from President Bill Clinton, and even was inducted into the National Inventors Hall of Fame in 2002.
In other words, he's a big deal. But alongside his many practical achievements, Kurzweil also has a long track record predicting future technologies.
Most notably, in 2005, he published the book The Singularity is Near, which forecast the future of AI, addressing the rapid advancement of computers, and the way humans would become reliant on technology and AI. It was, let's face it, worryingly accurate.
But what does the next 20 years have in store? Kurzweil is now answering just that question in follow-up book Singularity is Nearer, a collection of predictions in how AI will have an even greater influence on our lives.
We sat down with the man himself to unpack the biggest changes to expect.
AI will master the art of being human
One of Kurzweil’s key predictions revolves around Artificial General Intelligence (AGI). While many AIs at the moment are designed for a specific task, an AGI would match or surpasses human capabilities across a wide range of cognitive tasks.
Currently, while impressive, AI models are only able to work in one area. ChatGPT is a text-based model, Dall-E or Midjourney for images, and companies like Wix are using AI for web design.
Artificial General Intelligence is a hypothetical software that can do it all – it could learn and adapt to new skills and situations and understand human reason. Think Scarlett Johansson's virtual assistant in 2013 film Her.
According to Kurzweil, an AGI capable of this will be available by 2029. “But that’s now starting to look like a conservative view," he tells BBC Science Focus. "Other experts say it will be two years, maybe three."
He adds: “When I first said this date in 1999, people found it alarming. Stanford organised a conference of 1000 people to discuss how realistic this was, and their view was that it would happen, but not within 30 years – the estimate was 100 years.”
While 2029 will feel unrealistic to some, it follows the speed at which artificial intelligence has risen. Kurzweil points to the exponential gains seen within the technology, which are growing faster each year.
“Economists assume that the flow of these technologies is linear – it goes 1, 2, 3, 4. But really it is more like 1, 2, 4, 8. When something grows that quickly, advancements seem to start happening so suddenly one after the other,” says Kurzweil.
“Today’s computers can do half a trillion calculations per second. That would have been seen as an impossibility just 10 years ago.”
The end of ageing
One of tech’s big talking points right now is a rather lofty aim – to defy ageing. While many have tried, Kurzweil believes we are nearing a point in time where we can not just slow ageing, but combat it entirely.
“We’re going to solve ageing in the next 5 to 10 years. Right now, as you get older, the probability of you dying the next year goes up. By the time you get to your 90s or older, there’s a very high chance of dying each year,” says Kurzweil.
“We’re going to overcome that. It’s already happening in medicine – we’re seeing AI rapidly speed up medicine and drug discovery. By the time we get to around 2029 to 2035, I firmly believe we will reach longevity escape velocity.”
A somewhat controversial concept, the longevity escape velocity suggests that people can live indefinitely by extending their remaining life expectancy faster than time passes. To achieve this, technology would need to advance to the point where it could actively achieve medical intervention equal to life lived – effectively gaining a year for every year lost.
If this sounds too good to be true, you’re not alone. This theory is hypothetical and faces a lot of blockades. Ageing is a highly complicated system, made up of a variety of factors within and outside the body.
However, people are already living longer than they did just a few years ago. With rapid expansions in health and technology, the average life expectancy is likely to increase.
“This doesn’t guarantee you’ll live forever. Any 20-year-old can have complications and could die tomorrow. However, the likeliness will decrease. Companies are making artificial lungs and kidneys, treatments are exponentially better and our understanding of diseases is improving,” says Kurzweil.
We will merge with AI in the near future
In both his books, Kurzweil refers to something known as ‘The Singularity’. A term borrowed from physics, the singularity refers to a hypothetical future point in time where technological growth becomes both uncontrollable and irreversible.
Like many of his other predictions, Kurzweil puts a date on this: 2045. “This will be the singularity – where we no longer have control of AI. In physics, the term singularity means something so powerful that it exceeds our understanding so much that we can’t even imagine what will happen,” he says.
The so-called ‘Singularity’ is delved into deeply in both of his books, but there are a few key parts of the idea. The most noticeable of which is his belief that we will ‘merge’ with AI, creating a new form of intelligence.
This would mean a dramatically enhanced level of human intelligence, allowing us to overcome limitations. He also believes that nanotechnology will play a crucial role in connecting the human brain with computers, creating a seamless interface between the two.
Like many of the topics he touches on, Kurzweil’s beliefs around the singularity aren’t without criticism. It relies on the continued rapid advancement of technology and would require a far better understanding of both intelligence and human's ability to merge with technology.
One thing is for sure, Kurzweil is excited about the future, ready to defy ageing and merge with AI by 2050.
About our expert, Ray Kurzweil
Ray Kurzweil is an inventor, author and futurist. He has worked in the field of AI for 61 years. In 2005, he published the book The Singularity is Near. This addressed the future of AI, establishing a variety of commonly held believes about AI today. He followed this book up in 2024 with the book The Singularity is Nearer.
Read more:
