The United States can no longer afford to risk so many lives and airplanes against increasingly lethal Soviet defenses and must put more emphasis on unmanned weapons, the Defense Department's research director said yesterday.

Donald A. Hicks, who became Pentagon research chief last month, said he will recommend significant increases for such "smart" weapons, starting with at least a 10 percent increase in the coming year. He cited the loss in Lebanon of two Navy bombers and one pilot to Soviet missiles fired by Syrian gunners in a December 1983 bombing raid as a situation where smart weapons could have spared men and machines.

"We have only begun to scratch the surface in smart, standoff weapons," Hicks said in an interview.

Navy officials said the targets they were ordered to bomb in Lebanon were too small to allow use of smart weapons. Pentagon officials, asked about this, said the experience suggests the need for spotter planes that could project a laser beam that bombs would follow to the target.

With technology now available, the United States could assemble a pod of cluster bombs that could be released by an airplane 20 to 40 miles from the target. The pod would then fly itself over the target and automatically drop the bombs, Hicks said.

Hicks and other Pentagon executives, including several recently recruited from the aerospace industry, will try to convince the services to put more faith in smart weapons during the second half of President Reagan's rearmament program. Military leaders, who often are disappointed by modern weapons, traditionally are reluctant to switch from the proven to the unproven.

The Army, according to critics inside and outside the Pentagon, refused to recognize how lethal smart Soviet missiles had become as the service developed its Divad air defense gun, which consisted of two fast-firing 40-mm cannons. Defense Secretary Caspar W. Weinberger canceled the Divad last week -- after the Army had spent $1.8 billion -- because it was vulnerable to Soviet missiles launched from helicopters beyond the range of the guns, Pentagon officials said.

Hicks plans to meet with Army leaders on Friday to discuss possible substitutes for Divad as new countermeasures are developed to Soviet smart weapons.

Although the United States has developed some relatively simple smart weapons, they have not been widely distributed to the troops. The Soviets have been remarkably effective at copying those U.S. weapons and getting them to the field quickly, according to defense officials and outside specialists.

Under the tight U.S. military budgets likely in the future, the officials added, considerable sums will have to be spent on weapons that can survive against those smart Soviet missiles.

Former defense secretary Harold Brown, while leveling no criticism at the team now running the Pentagon, said defense planners must reckon with this new reality of battle: "If you can see them, you can kill them."

Smart weapons guide themselves to their targets by following maps stored in their mechanical brains, by feeling their way with radar, by homing in on heat, or by riding beams of light. Specialists said one set of challenges now involves how to confuse, dodge, destroy or hide from these precision weapons.

William J. Perry, former Pentagon research director, said in an interview that the Army's slowness in adopting smart anti-armor weapons and related technologies "has been a fiasco."

"I have been discouraged by how long it has taken to get smart weapons into the field," Perry said. "Even if Divad had acted exactly as it was supposed to it was outranged by standoff weapons. This left as its only justification a cheap way of doing a limited job. But it did not end up being cheap."

The cost would have been about $5 billion for 618 Divads, about $8 million each.

Retired Lt. Gen. Kelly H. Burke, Air Force research and development chief from 1979 to 1982, said he -- like Perry -- is dismayed by how slow the military has been to embrace and deploy smart weapons while the Soviets copied U.S. technology and fielded it. He said there must be a recognition in planning both offensive and defensive systems of how lethal smart weapons would be in any next war.

Declaring there are bound to be heavy losses to smart weapons fired against aircraft, Burke said: "There may be such a thing as a cheap airplane, but there's no such thing as a cheap American pilot." Therefore, he said, it is time to start mass producing unmanned systems with the recognition that many of them would be destroyed in any war with a country armed with smart weapons.

"The dumbest thing in the world is to put a pilot in a sophisticated fighter-bomber and tell him to attack a fixed target," Burke said.

"We have to break our mind-set," Burke fretted. "We can't seem to just turn out a mass of unsophisticated weapons" like drones and other remotely piloted vehicles, called RPVs. "Maybe we should just go to the auto industry and say, 'Here, build us these to automobile standards.' "

"I'm worried about ever more smaller numbers of ever more capable systems," Burke said. "Quantity has a quality of its own."