Jack Northrop believed the flying wing was the most aerodynamically efficient shape an aircraft could take: no fuselage, no tail, nothing but lift-generating wing. In the 1940s he built the piston YB-35 and the jet YB-49 to prove it, but the aircraft were unstable and hard to fly, and the programme was cancelled in 1949. The story goes that Northrop, by then frail and near the end of his life, was quietly shown a model of the still-classified B-2 in 1980 and understood at last that his idea would fly.

What changed between the YB-49 and the B-2 was electronics. A tailless flying wing is inherently unstable, constantly trying to wander in yaw. The B-2 is kept pointing straight by a quadruple-redundant digital fly-by-wire system making tiny corrections thousands of times a second, using split “deceleron” surfaces at the wingtips for both drag and steering. The shape Jack Northrop chose for efficiency turned out to be almost perfect for stealth as well — a happy accident that made the flying wing the ideal form for a bomber meant to vanish from radar.

Stealth is not one trick but several working together. The B-2’s shape does most of the work — every edge on the aircraft is aligned to a handful of angles so that radar energy is scattered away in a few narrow directions rather than back at the transmitter. On top of that, the airframe is coated and built with radar-absorbent materials that convert radar energy to heat, and the engine intakes and exhausts are hidden on the upper surface to mask the hot, radar-bright turbines. The price is maintenance: the special coatings historically needed climate-controlled hangars and painstaking care after every flight, one reason the B-2 has always been costly to keep mission-ready. Exact radar-cross-section figures remain classified, and any single number you see quoted should be treated with caution.

A tailless flying wing has no fin to keep it tracking straight, so left to itself the B-2 would yaw and wallow. It stays controllable only because a quadruple-redundant fly-by-wire flight-control system constantly nudges its control surfaces — including split deceleron panels at each wingtip that open like clamshells to create drag on one side and swing the nose. Its four General Electric F118 turbofans have no afterburners; the B-2 is a subsonic aircraft that trades speed for stealth and efficiency, cruising high and quiet rather than dashing. That combination of computerised stability and clean, buried engines is what lets a shape Jack Northrop could never quite fly become a dependable long-range bomber.

The B-2 is routinely called the most expensive aircraft ever built, but its price depends entirely on how you count. The pure flyaway cost of one airframe has been cited at roughly $737 million in 1997 dollars, and the Air Force has quoted around $1.16 billion per aircraft in fiscal-1998 terms. But because the enormous research-and-development bill was spread across just 21 aircraft after the fleet was slashed from a planned 132, the total programme cost per aircraft climbs above $2 billion — the figure most often quoted in headlines. All of these are legitimate; they simply answer different questions. Treat any single dollar figure for the B-2 as one accounting choice among several.