Interesting article, but there are clues that Betelgeuse is not
actually shrinking. The article states:

"Since the 1921 measurement, its size has been re-measured by many
different interferometer systems over a range of wavelengths where the
diameter measured varies by about 30 percent," Wishnow said. "At a
given wavelength, however, the star has not varied in size much beyond
the measurement uncertainties."

Note the wavelength-dependent 30% variation in size. The final
sentence of the above paragraph directly clashes with the thrust of
the article, that there is now indeed a shrinkage of 15% in the one
wavelength. This is suggestive that if Betelgeuse undergoes natural
internal cycles (of heat or whatever), that the wavelength-dependent
visible diameter may migrate to longer or shorter wavelengths. The
article continues:

"The measurements cannot be compared anyway, because the star's size
depends on the wavelength of light used to measure it, Townes said.
This is because the tenuous gas in the outer regions of the star emits
light as well as absorbs it, which makes it difficult to determine the
edge of the star."

The article then describes that Townes finds a way around this
problem, but with Betelgeuse now said to be shrinking by 15%, it is
apparent that Townes has not got the answer after all. I trust the
star more than the astronomer.

Eric