p.25 The test of Fritz 8, test version 15 took place in the summer of 2003 in Hamburg on an office computer
at the ChessBase offices. Programmer Mathias Feist was assisting, and interpreting Fritz's thoughts for me, while explaining
about the recent areas of focus in the programming of Fritz... When I told Mathias that I had the feeling that the recent
commercial version of Fritz had a much higher respect for dynamic features, a big wide smile broke out on his face. "Yes,"
he said. "We worked a lot on this. It is playing chess now, not just counting pieces. It took some time to come around to
it, as it is obviously much harder to evaluate dynamic features than it is to simply count material."
p.28 This is a fortress, a thing a computer is not built to understand, as the advantage never expires for
Black.
p.69 "When Fritz does not understand the position, it is because it does not have all the accessible
knowledge. This means that the evaluation function can be improved even more... Fritz does have some understanding
of compensation... but the problem is that you cannot set it too high. The difficult thing is to balance it... Fritz should
know about this... This is obviously a position Fritz does not understand... It is a position even I understand! Something
is wrong here... a lack of understanding."
p.97 [Mathias Feist, Fritz programmer] I understand very well that a weakness is only a weakness
if it can be attacked, but you cannot put this into the evaluation function. It is a matter of search... To exploit
them the program has to search. [JLJ - Perhaps there is another way to do this, see A Proposed heuristic...]
p.157 Chess has changed and we understand it better because of the computers. But that
does not mean that we can learn anything from the way computers think. Whenever someone tells me chess is only calculation
and uses computers as argument I laugh my (censored) off. The day we can calculate 2 million moves per second this
argument will be correct. But hardly before then. Still, with a few moves per second the best humans are doing well
against the computers. And the reason is that we think differently from them. Obviously tactics
are mainly concrete and should not be understood differently. But chess is not only calculation and intuition. It
is also logic, understanding of where the pieces belong, and long-term strategy. Here humans still have a
great advantage over computers. [JLJ - actual quote, the (censored) is reproduced as written. Certain positions
in this book are analyzed with Fritz 8, so this perspective must be understood when Aagaard refers to the "computer". ]
