Reactor race: did we back a loser? Atomic Energy: A New Start, by David E. Lilienthal. New York: Harper & Row. $8 .95.

The important message contained in this small book can be summarized very simply: David Lilienthal, the first chairman of the Atomic Energy Commission and the foremost early proponent of using nuclear fission to generate electricity, says that nuclear power, today in America, is unsafe. He is willing to accept part of the blame. He warns that we must deal with the dangers quickly and drstically. Pass it on.

Seventy nuclear power reactors are at present licensed to operate in this country, and more than 90 others stand at various stages of development. If you live in the Northeast, as Lilienthal notes, about one- third of the electricity used in your region comes from splitting atoms; in Chicago, half the electricity is nuclear, with a pair of reactors sitting 40 miles outside town; and if your home is Harrisburg, Pa., you don't have to be told about Three Mile Island. Of those 70 reactors now functioning, most have been manufactured by just four corporations (Westinghouse, Babcock & Wilcox, General Electric, and Combustion Engineering), according to a general principle of design that is known as "light water" technology. "Today the light water reactor, with its dangerous side effects," says Lilienthal, "is virtually the only nuclear generation facility available to the electric utility industry."

And not just in the United States. He adds that "we have made this particular technical system -- the light water reactor -- dominant in the international energy market. Now, belatedly, we recognize that our nuclear technology is not really so advanced. It isn't dependable enough, it isn't safe enough." Lilienthal is not wholly against the idea of using fission to produce electricity; but he believes passionately that, very early on, in choosing which technical method to pursue and subsidize, he himself and a few other government and industry officials made a very grievous mistake.

"I do not want to minimize my own role in this inadequately considered rush to adopt the light water method of atomic energy production," he avows manfully. That adoption of light water was partly a consequence, he explains, of a change in policy that allowed independent venturing in the field of atomic energy by private corporations, which during earlier days had served only as contractors for government work.

The first two American reactors to produce electricity were both light water types, built by General Electric and Westinghouse, according to two differing design variations that those companies respectively favored. One result of such private involvement, Lilienthal says, "was to set the stage for the subsequent commitment to the light water fission conversion method, a method whose defects we were aware of but which, because of ther inexperience with a new and complex technology, and in our haste, we failed to take into full and sobering account."

Another contributing factor was the influence of the Joint Congressional Committee on Atomic Energy, which attained almost executive authority in its area, and through the 1960s prompted enthusiasm and corporate investment on behalf of light water technology. Words of caution from a few experts did not dampen the boosterism, and orders stacked up for light water reactors. "Among the companies that benefited the most," Lilienthal notes, "was the designer and manufacturer of The Three Mile Island reactor, Babcock & Wilcox."

Now, 35 years after it all began, the industrial world has spent $200 billion trying to find a comfortable means of getting electricity from the atom, and the globe is peppered with these machines that he warns us against, with more still on the way. So what is to be done?

Lilienthal proposes the "new start" of his title: turn back from the blind alley, renounce the vested financial interests, and begin over, chartering a New Atom Corporation to conduct fresh research and encourage diversity of approach. But how to inoculate his New Atom Corporation against the mistakes of his old AEC? How to convince those huge private vendors to abandon their wrong-headed investments? Lilienthal doesn't say. Or at least, in this book he hasn't.

Please, Mr. Lillienthal; tell us how.