Machine intelligence will catch up with that of humans and begin to overtake it in the next two decades, a visionary scientist has predicted.

Inventor, author and futurist Ray Kurzweil forecast the fields of computing and biology coming together to take artificial intelligence(AI) to undreamed-of new horizons.

Much of the progress needed would be achieved by "reverse engineering" the human brain -- a process that was already well under way, Dr Kurzweil told the annual meeting of the American Association for the Advancement of Science in Boston.

"Three-dimensional molecular computing will provide the hardware for human-level 'strong AI' by the 2020s," he said.

Already, two dozen regions of the human brain had been modelled and simulated.

Dr Kurzweil added: "Once non-biological intelligence matches the range and subtlety of human intelligence, it will necessarily soar past it because of the continuing acceleration of information-based technologies, as well as the ability of machines to instantly share their knowledge."

Intelligent "nano-robots" -- microscopic units of AI hardware -- would eventually become "deeply integrated" in the environment, our bodies and brains, he said.

Human intelligence would be enhanced, and life expectancy vastly extended. "Full immersion" in virtual reality worlds would also be a possibility.

Dr Kurzweil, who pioneered speech recognition technology and has written several "future-gazing" books, was one of 18 "maverick thinkers" invited by the US National Academy of Engineering to identify the greatest technological research challenges of the next 100 years.

Others included Google co-founder Larry Page, geneticist Dr Craig Venter, and climate change expert Professor Rob Socolow.

The 14 "grand challenges" listed by the panel were:

• Engineering better medicines

• Advancing health informatics

• Providing access to clean water

• Providing energy from nuclear fusion

• Making solar energy economical

• Restoring and improving urban infrastructure

• Enhancing virtual reality

• Reverse engineering the brain

• Exploring natural frontiers

• Advancing personalised learning

• Developing carbon sequestration methods

• Managing the nitrogen cycle

• Securing cyberspace

• Preventing nuclear terror

Prof Socolow, from Princeton University, New Jersey, said: "In some respects we were given a crazy assignment: to identify and rank the greatest engineering challenges that lie ahead in the 21st century.

"Ultimately we didn't find it within our intellectual powers to rank these challenges. How do you rank the eradication of poverty versus keeping the planet habitable versus avoiding nuclear war?

"Instead we came up with broad categories of the challenges that lie ahead and within those categories, identified specific initiatives."

© 2008 Australian Associated Press Pty Limited (AAP) or its Licensors. This is the ZDNet Australia TM service with content provided by AAP. AAP reserves all rights, including copyright, in services provided by it. The information, text and images in the service are for personal use only and may not be re-written, copied, re-sold or re-distributed, framed, linked or otherwise used whether for compensation of any kind or not, without the prior written permission of AAP.
This service is published for information only without assuming a duty of care. AAP is not in the business of providing professional advice, and gives no warranty, guarantee or other representation about the accuracy of the information or images contained in this service. AAP is not liable for errors, omissions in, delays or interruptions to or cessation of the services through negligence or otherwise.
The globe symbol and "AAP" are registered trade marks.