We believe that the future 
of algorithmic information dynamics

and model driven approaches are bright,

and we would like you to be instrumental

in this change of vision
towards this science

based on models and computation.

We believe that the way forward is

to complement statistical approaches
and current approaches

with better algorithmic tools and methods.

There are many challenges in the area.

Some of them involve
finding ways to connect

the discrete nature of computation

and the continuous nature
of data and other areas,

such as differentiable programming
and deep learning

that are currently heavily based on
continuous functions,

but also with other
model driven approaches,

such as differential equations.

One first major milestone achieved

is the numerical calculation
of algorithmic probability by BDM,

and showing its stability and usefulness,

because our approaches produce
actually real continuous values

from an empirical distribution

in contrast to integer values,
as, for example,

from using lossless compression algorithm
on Shannon entropy

in the estimation of randomness
and algorithmic complexity.

So, one can use BDM
to replace the cos function,

for example in deep neural networks

and still keep it differentiable,
in principle.

But most work in this direction
is yet to be done,

and we have projects in which
we think you could get involved

if you wish, as we speak.

To tackle these tasks
and exciting new areas of research,

we are creating working groups
to which you can get associated.

There will be team leaders that

will be guiding the various
research groups,

divided by research projects.

Among those leaders will be
likely Narsis, Alyssa and I,

together with some other brilliant
collaborators in different areas,

from our collaborating network.

Each project will aim at
publishing one or more papers.

And I will be hoping to
consider papers

for the journal 'Complex Systems'

assuming I'm not listed as an author.

And, if I'm listed as an author,

assuming I have contributed enough,

then I'm sure we will find
the best journal for the paper.

If you're interested in getting involved,

join the online group on your screen.

You can even work on your own,
as some people prefer.

And of course,
you don't need to join any group

to conduct research on your own

on algorithmic information dynamics
or using our tools.

Finally, I want to warmly thank you,

all of you, to have followed our course.

We know there is a long room
for improvement

and this course will continue running

and hopefully getting better
in the next years.

But you have been the first generation

and we are extremely grateful for it.

Our algodyne team thanks you again.

Thank you very much!