As AT&T Corp. has refashioned itself for a world consumed by the Internet, spending $120 billion to seize the nation's largest collection of cable television franchises, many in the industry have looked on in awe. The company plans to turn cable into the pipes of a new communications experience while delivering consumers long-promised local telephone competition. Chief executive C. Michael Armstrong has been widely hailed as a visionary.

But as the Harley-Davidson riding Armstrong prepares to command a stage in a hotel ballroom Monday to address Wall Street analysts, the hosannas have been eclipsed by gnawing worry. One of America's signature brand names is irretrievably bent on a bold course, reaching deep into unexplored territory. If the promises are rich, so too are the questions.

Did AT&T pay too much for its cable properties? Can it really turn Tele-Communications Inc., a primitive amalgam of cable systems, into the swift pipes of the next communications age? What will it all cost? And when will the profit materialize?

"Their challenges are the most significant that any major corporation faces in America," said Brian Adamik, a telecommunications analyst at Yankee Group in Boston. "The strategy's great. It's execution that matters. And AT&T has historically been a poor executor. Can they overcome that? I give the odds at fifty-fifty."

More than AT&T's success may hang in the balance. The company--and cable in general--has emerged as a key candidate to deliver the local telephone competition that was supposed to accompany the deregulation of telephone markets. AT&T bought cable because the wires reach into homes, supplying the means to get around local phone companies. If cable companies deliver high-speed Internet access, the argument goes, local phone companies will be forced to roll out their own rapid technology, DSL.

"In cities where cable companies deploy, it lights a fire under the local telephone companies," said Federal Communications Commission Chairman William E. Kennard, who has declined to regulate Internet access over cable lest AT&T and others lose the drive to upgrade their systems. "It's about consumers getting choice."

Industry executives say Armstrong had no choice but to buy into cable. AT&T's only other option was to cling to its eroding perch as the nation's largest long-distance company.

The company still claims about half the nation's long-distance telephone service, which generates about 55 percent of AT&T's revenue. But because of price wars and new competitors, AT&T is losing about 5 percent of the residential long-distance market a year.

Many of AT&T's customers trace back to the days before federal regulators broke up Bell Telephone. As many as half have never switched long-distance carriers, said a marketing executive at another telecommunications company.

In some ways, that's good. Most such customers are not enrolled in discount plans and thus pay high rates. But they also are less likely to make many phone calls. The ones that do are increasingly realizing those nickel-per-minute television ads and telemarketers might be talking to them after all. Many are abandoning AT&T. Others are staying, but switching to discount plans. Factor in the looming entry of local Bell companies to the long-distance business and AT&T's future appears more precarious still.

Armstrong knows how quickly dominance can unravel. He spent three decades at IBM, watching the computer pioneer fail to grasp the emergence of software and personal computers, losing its place at the top. More recently, as chief executive of Hughes Electronics Corp., he oversaw the roll-out of DirecTV, the satellite television service that now steals customers from his cable franchises. He sees long-distance as no safe haven for AT&T.

"We were going to sit around and defend that 48 percent of the long-distance market, or we were going to go on the offensive," Armstrong told a conference in New York last fall.

AT&T officials say the company is on track to upgrade half of TCI's cable systems by the end of the year. The cable unit is on target to claim as many as 200,000 Internet customers by year's end, with 1 million as the goal next year.

"You're going to start to see scale being created and delivered next year," said Daniel Somers, AT&T's chief financial officer, who is acting head of its cable and high-speed Internet unit in Denver. "I expect to produce solid revenue growth. . . . No one says it's going to be easy to do what we're going to do, but we will do it."

The company hopes to soon gain regulatory approval for its other cable purchase, Media One Group. The FCC is digesting a letter from AT&T that argues the purchase would not violate laws limiting the allowable cable holdings. At the center of the debate: Whether AT&T's 25 percent stake in a partnership with Time Warner would give the company undue influence over programming. AT&T says it has no influence. FCC officials say they aren't sure.

Many are bullish. Last week, Jack B. Grubman of Salomon Smith Barney, one of the most influential industry analysts and long concerned about the cable venture's challenges, issued a report recommending AT&T's stock, finding "increased comfort level with the economics and the technology." Recent word that AT&T will issue a new stock to track the value of its high-flying wireless holdings has buoyed the company's stock, reversing several months of decline.

But questions persist. In this era of low unemployment, finding technicians to upgrade TCI's systems has been troublesome. Technical glitches have occasionally bedeviled its Internet service.

"Getting qualified people is a significant issue for us," acknowledged Carl Vogel, AT&T's chief operating officer for field operations at its cable unit, during a recent media briefing in Denver. "This is a product that we have no problem selling. It's a product that sometimes we have a problem installing and provisioning."

A cable industry executive with knowledge of AT&T's experience says the company is overstating its prospects and underemphasizing the pitfalls. "Mike paid too much," the executive said. "They don't have a lot of customers. They had hoped '99 was about a lot of progress operationally and a lot of deals. There's no deals and there's no progress."

Deals may be key. Even if AT&T claims Media One, it will still have links to fewer than a third of American homes. The cap on cable ownership prevents it from buying more. AT&T has counted on striking deals to sell telephone service over other cable systems, but none are in hand and talks are cold, sources close to the company say.

Prospects suffered a blow in October with the departure of Leo Hindery, the company's top cable executive, sources said. Hindery enjoys rapport with other cable executives, having run TCI, but Armstrong has poisoned relations by discussing a deal with America Online Inc., the nation's largest Internet service provider, sources say. Dulles-based AOL is vilified in the cable world for leading an ongoing battle to force cable companies to share their lines with Internet providers.

Somers, Hindery's acting replacement, remains publicly optimistic deals will be struck. "Our relationships with the cable industry are pretty solid," he said. "As we rock and roll here, we'll gain more and more allies."

At Monday's meeting, AT&T will formally unveil a new "fixed wireless" initiative, to direct telephone and Internet connections into homes and businesses via roof-top antennas. The plan will be presented as a way to broaden AT&T's reach to homes that are not touched by cable, one made possible as costs have dropped close to those of installing cable service.

But such claims provoke doubt. AT&T plans to use the same spectrum that now carries wireless telephone calls. A telecommunications engineer with a rival company said that the amount of spectrum AT&T owns would not allow Internet connections at near the speeds of cable or DSL. Some analysts and industry executives say AT&T knows its fixed wireless plan won't work: Its announcement is merely a bluff designed to frighten other cable companies into doing deals.

But Laurence Seifert, AT&T senior vice president for wireless, who heads the operation, said the venture is real, and tests in Dallas have already shown it will work.

"I'm spending a lot of money," Seifert said, declining to provide details. "We're building out base stations all across the country. I don't think AT&T shareholders would be too happy if we spent that sort of money on a fake-out."

For those taking the long view, the doubts and missteps are predictable.

"If it takes 200 steps to implement something, and you complete step 39, all you do is go on to step 40. No one cheers over that," said Rex G. Mitchell, an analyst with Banc of America Securities. "But if you fail on one of those steps, people like me are going to magnify that, and say, 'Oh, I wonder if that's a problem they're going to have on the whole system.' "

So far, he added, signs say the mission will eventually bring real profit.

CAPTION: Harley-Davidson riding AT&T chief executive C. Michael Armstrong faces a bumpy road ahead.