Hi community. I made my game programming portfolio at http://nicolasbertoa.wordpress.com/ I developed 2 games: - A 2D space shooter made with SDL - A 3D FPS made with Irrlicht engine and Irrwizard framework. I finished the following books: C++: - Thinking in C++ Vol. I and II - The Design and Evolution of C++ (Stroustrup) - Effective C++ (I don't finished More Effective C++ yet) - Professional C++ Math: - Essential Mathematics for Games and Interactive Applications, Second Edition: A Programmer's Guide. OO and Patterns: - Design Patterns Explained: A New Perspective on Object-Oriented Design (2nd Edition) (I dont finished yet) - Object-Oriented Design Heuristics API: - Introduction to 3D Game Programming with DirectX 9.0 (Wordware Game and Graphics Library) AI: - Programming AI By Example (I dont finished yet) I will try to explain what I want. I want two things: (1) I want to be an excellent OO designer and an excellent C++ programmer. What am I doing for that? I readed many C++ books. I'm reading "Design Patterns Explained" and I find that patterns are great! What will I do? I will read "Refactoring" and "Modern C++ Design". (2) I want to specialize in the AI field in games: What am I doing for that? I'm reading "Programming AI By Example". What will I do? Here is the problem. I want to become an expert on a specific AI topic, but I don't know the important topics.Then I will learn about the topic and I will develop an application(like a little game) showing what I learned. What stuff do I need to put in my portfolio if I want to get a job in the AI programming? Thanks!

1-reading won't get you anywhere. When looking at a portfolio people don't really care what you have read, only what have you done.
So for start, start making AI programs, lots and lots of AI programs, for games, for business applications and anything else you can thing of that requires inteligence.

so instead of having "I will read books" int "what I'll do about that" you should have "I'll make a lot of programs based on things I've learned and then try to create my own Ideas".

As already said,reading is not enough, you need demos/examples of working AI code you have created. There're several field of interest in AI, here are just a short list:
- pathfinding
- navigation mesh
- waypoint system
- decision making
- behaviour
- state machines
- behaviour trees
- steering
- flocking
- natural movement
- neural networks

Code one or more demos containing the above topics to prove that you're able to do basic AI coding. Present demos and videos of them on your webpage.

Then try to make demos of more advanced AI topics like collision avoidance in a crowd, npc looking for cover depending on the player position, a working simulation of small village, whatever. Best to look for current 'state-of-the-art' games which are praised for their AI and try to create something similiar or even better !


--
Ashaman

(You haven't read my book? I'm crushed!)

If I were you, and really wanted to break in to the game industry, I would start looking at where some of the newer games fall short in the AI department and then try to improve on them.

Even if you don't succeed you'll at least be able to demonstrate that your experience is relevant in todays market.


For example:
- Fallout 3 has some pretty complex environments, and I'm sure Bethesda spent a lot of time on path finding. It's very well done but still acts a little funny sometimes.

- I read somewhere that Fight Night Round 4 uses pattern prediction to help the computer box, but IMHO it falls short in the AI department as the computer has to 'cheat' in order to be competitive (the CPU does things you can't)

- Some of the Poker games I've played on my phone use frustratingly cautious CPU players-- to they point that they are easy to kill by death from a million cuts, but not fun at all to play against.

@Ashaman73 and spliter:

Yes, I need to make more demos showing what I learned.


@InnocuousFox:

Jajaja, when I finish Programming AI By Example, I will read your book, I promise :)