It has turned a bit into a slog.
Technically, there is one week of coding left to complete.
But the thing is that most of the work is of a conceptual nature, to figure out how intelligence truly works. It's big and the progress that has been made is mind-boggling. The fundamental connections between quantum physics, how the world organizes itself, down to the very fabric of space and time, and in the middle of all the conundrum of life, which is intelligent.
It's difficult to see, how this is relevant to develop the AI for H5. It's now a good time since I solved the biggest challenge how to move from processing innumerable variants (40 turns lookahead for 8 factions with multiple heroes interacting with each other and countless map objects) to a solution that understands what matters. The solution is solid and good, outstanding even, but I am still climbing that tower. It's like a spiral stairway, you arrive at a solution and then it reveals how you can go even further, again and again, and you keep climbing the tower.
It's now near the point at which I will simply take what has been achieved and write the final bit of code.
It's almost seven years since I started working on the AI for good, ceaselessly, day for day without a break. Well, there were a few exceptions, the day I met the classmates of my final year at school, and I think there were three or four days I were sick, and Wednesdays is mostly doing other chores. Somehow it's doable, and it's possibly the only way to develop true AI in one's lifetime. I cannot imagine that a work like this can be done in a regular development studio with roadmaps and schedules.
But in essence, it's mostly done now, and in theory you might get a new H5 to play in a couple of weeks, but please do not expect that.
Wait and see and hope for the best. And naturally there is a blueprint for a spiritual successor of Heroes of Might & Magic that has come along nicely, and which I believe most of you will like, even if the setting will surprise you.