Had an interesting day. I have had unity installed since last year, but not followed the tutorials as waited for the lecture to start. I loaded the karting simulation which on my mac took along time. Once it started, all the fonts were corrupted, so i could not follow the tutorial.
I then loaded NetLogo, and followed the tutorials – i have to say i was really amazed at what this application could do with such “basic” code. Having just completed the python/turtle assignment in the last semster, the code which produced AI, albeit basic, was amazing. Very high level and easy to read (reminded me of BBC B Enhanced BASIC language).
I tried the demos, first the tutorial, the wolf/sheep simulation, getting anquiated with the various methods to stop and start a simulation and the button types. I really enjoyed the wolf/sheep/grass simulation and once I had completed the tutorial and looked at the code, started to look at other routines in the library.
As we are doing gaming and AI, i loaded ‘pacman’, and whilst basic, it was a really enjoyable game ! I tested it thoroughly for all the features. I was amazed at the code, at 342 lines, could produce such good results, not just the game, but the AI in the game, how the ghosts ‘looked’ for the pacman, and once ‘energized’ ran away, the code was so simple, but produced great results ! I am quite tempted to use NetLogo, but have ordered a more powerful mac to help me with both my professional work and my academic studies as my current mac (Late 2013 model) is really showing its age.
I also looked at the fractal routines, having run similar code in Python and was impressed how quick the mandlebrot fractal and UI worked so well. It was far faster than the Python code, which told me the python interpreter for iOS must be very inniefficent, as the code in function for Python was far more complex/detailed than the high level code (lograthamic/expotential) than what was used in NetLogo.
I played for one hour on the tutorials and have really enjoyed NetLogo. My new mac arrives by Monday and I will load Unity onto it, but I know that working with Net Logo will also produce really good results. The AI is very simple to interface, and powerful in results.
I do not have much time to play games, but do own a Playstation 4, mostly i used it for Netflix, Amazon Prime and Disney+, but as I’m doing Gaming AI wanted a game to play. I looked at games which had good AI here and after reading the descriptions and forming a short-list went onto research “Alien:Isolation”. The Wikipedia post convinced me, as a stand-alone 1st person AI game, it was the right one for me. I dont have time for multiplayer games, and the reviews for “Alien: Isolation” really stood out. I ordered the game and a headset, as the AI picks up on sound and wanted to experience the full emersiive experience ! I look forward to trying it out, but I dont think my wife will be watching whilst i play, unlike our Korean films on Netflix…
In our taught lesson I enjoyed the week 2 learning, the key points of AI in as various was really was really an eye opener. Opponent/Team Mate/Director/Scenery and Architecht, there is ALOT of what AI is doing in which we might otherwise not notice or just take for granted. One thing I do like about my ‘age’ is having experienced games from the 1970’s all the way to now and being able to reference this different types. Even on an Atari 2600, there is some degree of all of those AI features, albeit at a rudementray level compared to the immersive features in todays gameplay.
The part on ‘spawning’ was of very interesting and well articulated into terms i could understand. Having played games where most advesorys ‘spawned’ from the same point, this was interesting breakdown of modern games. I’m looking forward to experiencing this first hand in the ‘alien’ game. Where the alien cannot be defeated, so ‘spawning’ will be interesting in this game.
I continue to think about my own game, based on WSTJX and how i can model that. I first need to ‘render’ a world, quite literraly of the world broken into Maidenhead grid locators. I know it will be possible to make a global map in both Unity and NetLogo, but i think the visual experience will be richer in Unity, but from a computational basis, am confident that NetLogo can bring the game alive also. The section on PCG (Procedural Content Generation) gave me inisights on how i might render my ‘world’ and the things i put into it.
I ordered another books to help with my AI Studies, which come from the NetLogo tutorial.
- An Introduction to Agent-Based Modeling: Modeling Natural, Social, and Engineered Complex Systems with NETLogo (The MIT Press)
I also reached out to my Machine Learning / AI teacher from semester 1, whilst i had to leave the module due to work load, i do have the book she published (Pattern Recognition and Neural Networks, Ludmila Kuncheva) as I think the AI/ML concepts she was teaching will nicely augment what we are learning in this module.