This really makes me want to go back to my game and play some more with AI. Although, at the same time I realize that's the mistake I made - making the game/platform for creating AI. It really takes a toll. This guy uses existing NES games, which allows him to focus on the AI and even generic game learning. Brilliant.
You can clearly see the advantages of AI in the short-term play (think in the order of milliseconds, exact frame-by-frame button presses) over anything a human could ever achieve.
Imagine combining the short-term button-mashing of said AI with the long-term planning of human players...
Feed your bot AI with information that is 200 ms old and force it to extrapolate 200 ms in the future for its aiming/movement/etc. logic. It will make the bot much more human-like with noticeable reaction time, so it could be tricked by sudden changes in your behaviour. Such a simple but powerful AI trick.
Some of the tool-assisted speedruns feed scripts into the emulator. One of the pokemon yellow arbitrary code execution runs used some clojure: http://tasvideos.org/3767S.html
Both the video and paper are well worth going through!
The Abstract is 16 words. The introduction begins with
"The Nintendo Entertainment System is probably the
best video game console, citation not needed." Need we say more?
Also, I wish non-programming fields of media would also adopt pythonic triple-quotation marks.
They aren't "inch marks". They're straight double quote marks. The prime (used for feet, arc minutes), double prime (used for inches, arc seconds) and ditto mark are all distinct.
The paper is worth a full read as well. One of my favorite bits... "The FCEUX emulator is about a jillion lines of C++-ish code, was intended as an interactive GUI application, contains support for multiple di erent platforms, and the code is, on a scale from a pile of horse shit to not horse shit, approximately a 2"
As one of the major contributors to FCEUX, I feel that the author is misrepresenting the code quality of FCEUX. To be fair, I'd say the code would rate about 1.5 or maybe 1 on said scale.
That's just normal in Super Mario Brothers 1, kills by shell start at 500 points and escalate for more targets. All the later games start at 100 instead, SMB3 and SMW and the New SMB series.