You don't exactly have to push any buttons to set Michio Kaku off rehearsing the ideas in his book, Visions: How Science Will Revolutionise the 21st Century and Beyond, a breathless scattershot of outlandish facts, figures and predictions. Kaku has a fair handle on the scientific principles of the biomolecular, quantum physics and computing revolutions, yet he talks and writes in a wild mixture of the utterly credible and the collegiately deranged, couched in the kind of hyperbole beloved of a US audience: your entire genetic code on a credit card, shot-in-the-arm cures for cancers, halting the ageing process, starships and extra-terrestrial intelligence, etc.

A professor of theoretical physics at the City College of New York, Kaku has become famous since his 1994 bestseller Hyperspace. Visions is the product of five years hosting a US-wide radio science show, interviewing hundreds of leading scientists. "I got to visit all these centres for DNA, cloning, cyberspace, robotics and quantum physics. These scientists are very happy to lay out the timetable until 2020, because billions of dollars of American tax-payers' money depend on it." A man who has been a consultant both for NASA and Star Trek - surely a paradigm only dreamt up in America - Kaku's lecture to the Trinity College Physics Society last Friday evening was an alarmingly dumbed-down affair, an unashamed sales pitch for the book, the technologies it describes, and ultimately, one feels, for research funding.

He was leafing through USA Today in the Shelbourne when I came across him: an incredibly spruce 52-year-old, with a grey-tinged mane framing the boyish face. Chewing sandwiches without interrupting the rapid techo-patter, he outlined the big changes up ahead: "The engine is Moore's law, which states that computing power doubles every 18 months - and that's held true for 50 years with very few deviations. If anything it's accelerating. But, by 2020, because chips cannot keep shrinking - as silicon cannot sustain molecular-sized transistors - we're going hit the expansion limit, and Silicon Valley will become a rust belt. At that point, we're going to need a new architecture for computing." He's even happier talking about gizmos: "Your glasses will become intelligent. In the future, you'll be able to attend an emergency meeting at the Home Office by video conference without ever leaving the beach. Even your clothes will become intelligent, so you will never die alone of a heart attack down a lonely street. Your clothes will monitor your heartbeat, call the ambulance, locate your position by GPS, and download your medical history. In Japan, they've already marketed a smart toilet, which alerts you if you've eaten too much sugar or fat."

So in this clear vision of 2020, we won't even be able to trust our own underwear. But if some of it sounds like sci-fi, a lot of it is on-track, such as the first phase of the Human Genome Project, a $3 billion crash programme to locate all human genes by 2005; or the Californian corporation which has grown large areas of human skin by cheating the cells' biological clock and allowing them to continue dividing, making them, in Kaku's word, "immortal". I tried to slow him down, arguing the complexity of human tissues. "Sure, this is not a fountain of youth. Different cells age at different rates. Brain cells hardly divide at all; they age by oxidisation - the rusting process. But in the future, we may have a combination of therapies, like a cocktail of enzymes related to age genes, three of which have been identified.

"Or as organs give out, we'll simply grow new ones. In the last three months, embryonic human stem cells were found for blood. So when children are born, we may just tap a few stem cells and keep them in a bottle as spare parts. In principle we should be able to grow unlimited quantities of blood, and the next major organ to be grown will be the liver; it's just a chunk of tissue. Kidneys are more complicated, but we think they may be growable, and at Harvard, they have long-terms to grow an arm." An arm. "Yeah, you use a porous plastic scaffolding, and seed it with baby stem cells, for muscles and skin, with pumps to keep the cells alive. And they grow inside the skeleton, until that disintegrates, leaving a perfect organ in its exact configuration. That's how heart valves were grown. Dr Walter Gilbert at Harvard reckons that in 50 years, every organ will be growable." Kaku's own specialist field is as a "co-founder of string field theory". The larger field of superstring theory is currently being vaunted as a "theory of everything", linking the equations of quantum mechanics (which describe the three fundamental forces of electromagnetism, and the strong and weak nuclear forces) with Einstein's theory of General Relativity and the other fundamental force of gravity.

Kaku says: "String theory was stumbled upon by accident in 1968. Two physicists in Geneva found that the beta function described very remarkably the scattering of pi-mesons. Later we discovered that the properties of the beta function were caused by vibrating strings, but nobody had a comprehensive field theory, as Maxwell and Faraday did with electro-magnetism. But now you have an equation which is one inch long that gives you all of string theory."

Yes, but it hasn't been solved. "Correct. In fact, we've found millions of partial solutions, but we want the solution. We want three generation of quarks, electrons, neutrinos, protons, we want the Big Bang, galaxies, supernovas, we want DNA and even life to come out of this equation." Although as a theory of creation it is not experimentally verifiable, string theory is gaining credence, although Kaku's book cites one Nobel prize-winner, Sheldon Glashow, who opposes it. "He has since shut up. All the recent tenured slots are going to string theorists. Harvard has now hired three - in his own department. Glashow once vowed to keep string theorists out of Harvard. That was a standing challenge, but he gave up."

Born in 1947 in California, Kaku's grandparents were Japanese, although his parents were born in California. "As teenagers, they went back to Japan for a Japanese upbringing, and came back just before the second World War, which was the wrong time - there was a lot of hysteria. They were locked up from 1942 to 1946, and their property was confiscated. We were dirt-poor, but there was nothing left to go back to, so we stayed in California." Kaku still remembers how, when he was eight, his elementary teacher announced that Einstein was dead. "Every newspaper showed his desk with the unfinished manuscript of his great unfinished unified field theory. So I thought, maybe I can try to finish it." While still at high school, Kaku's obsession led to his building an "atom-smasher. I got scrap metal, 22 miles of copper wire, 500 lb of transformer steel, and I built a 2.3 million eV betatron in the garage. Of course it blew out all the circuit bracing in the house."

However, he won a prize for it at the National Science Fair. "That's when I met Edward Teller, the father of the hydrogen bomb. That was my first connection to the atomic elite. Teller took an interest in me, and I got a scholarship to Harvard. Little did I know he was using the scholarship to groom young physicists to work in Livermore National Laboratories designing third-generation hydrogen bombs.

"Still, about 50 per cent of nuclear scientists are connected to the military. If you can't get a job at the university, you go to the Pentagon. My attitude is that the Cold War is over, and we don't need these bombs. So, I've been working with different disarmament groups. We organised a boycott of the Star Wars programme, by getting about 6,000 nuclear scientists to sign a pledge never to accept a dime from Star Wars research." "Since then, we physicists have realised that science has become rather distant from the consumer and indeed the tax-payer. In the old days, we simply went to Congress and just told them the Russians were doing it. Now, we have to make a direct appeal to the American people about the wonders of science." He maintains that public education is a core reason for his books. "I always remember how difficult it was as a kid to find anything reliable about science. What was published in the popular press was awful, with high-school teachers who don't know up from down. Even now, science education in the US is uniformly very bad."

Visions by Michio Kaku is published by Oxford University Press, £8.99 paperback in UK