\begin{slide*}
\centerline{\shadowbox{Optimistic discussion}}
\vskip 1cm
\begin{center}
{\footnotesize
\begin{minipage}{0.96\linewidth}
\begin{enumerate}
\item 
with few ingredients and one control parameter, we can describe a wide range
of seismic sequences.
\item
The $A$ value can be determined from
large scale deformation and the $B$ value can be compared with its
equivalent in earthquakes catalogs.
\item
The low value of the average stress
corresponding to the highest seismic regime could also be compared
with in-situ measurements.
\item
the phase diagram allow us to describe the system in term of critical state
and to study the seismogenic potential of the fault zone
through a set of trajectories that we can also call ``{\em attractor}''.
\item
this model shows that it is possible to extract some
seismic precursory phenomena for large event if $A<0.5$.
\end{enumerate}
\end{minipage}
}
\end{center}
\end{slide*}

\begin{slide*}
\centerline{\shadowbox{Pessimistic discussion}}
\vskip 1cm
\begin{center}
{\footnotesize
\begin{minipage}{0.96\linewidth}
\begin{enumerate}
\item
highest slip rate fault (i.e. high $\Delta E \Rightarrow A >> 10$)
do not always exhibit creep.
\item
the variation of the $b$-value can lead to opposite
conclusions .
\item
the acceleration of the seismic release is
not obvious.
\item 
phase diagram trajectories avoid different parts of the phase diagram and there
is still unexplored zones which can have their counterpart in real fault zones
(e.g. high density of micro-fractures coupled with low stress).
\item
for $0.5<A<2$ the system
is chaotic and no seismic precursors are observed. Under these conditions
prediction becomes impossible.
\end{enumerate}
\end{minipage}
}
\end{center}
\end{slide*}