Consider a scalar field theory in $AdS_{d+1}$ spacetime
\begin{equation} S= -\frac{1}{2}\int d^{d+1} x \sqrt{-g} (\partial_\mu \phi \partial_\nu \phi g^{\nu \mu} + m^2 \phi^2 ) \end{equation}
On quantization of this theory in an AdS metric, one finds that the solution can be written as
\begin{equation} f_{\omega l \vec{m}} (r,t,\Omega) = \psi_{\omega l}(r) e^{-i \omega t} Y_{l \vec{m}}(\Omega) \end{equation}
where $\psi_{\omega l}(r)$ turn out to be hypergeometric functions. The key result here is that $\omega$ gets quantized as follows
\begin{equation} \omega = \omega_{n l} = \Delta (=\frac{d}{2}+ \frac{1}{2} \sqrt{d^2 + 4m^2}) + l + 2n \end{equation}
And this quantization somehow implies that mass is not properly defined in the theory? I don't quite understand how this follows just from the fact that the $\omega$ is quantized?