Started in 2014, Project A.G.INT. is currently on its 3rd revision.
Essentially ended up as a processor instruction set.
A bunch of algorithms in areas such as text understanding / synthesis and
computer vision. Since the algorithms didn't allow
for adaptability, and were limited in the variety of scenarios they could
handle, it was realized that this approach was very far off from what the
human brain actually does. Due to this, work was scrapped and Project A.G.INT.
moved on to its second revision.
A higher level look at the mind. This version focused on the higher level
functionality of the different processing regions of the human brain, and also
included psychological models of the mind. Near the end of this version, parts
started getting too specific, similar to what happened with V.2. However, some
of the psychological observations were found useful, so while most of the work
was left behind, these observations were used while transitioning to V.4.
The current model, fully neural, the approach for which is described in the
about page.
V.4 was started on an investigation into blindsight, and what perception actually is.
