Lessons from the games industry part 2: Process AI and ACM evolution

My process can beat your process

Following on from previous posts on BPMredux (Social enterprise, the Cabal, and ambient awareness) and taking up on Scott Francis’ suggestion that more cross-pollination is needed in the BPM industry at large, I decided to scribble a few notes on Process AI and the industry starting to hire outside it’s own magic circle for innovation.

There was a recent article on ArsTechnica relating to a project where a computer creates a set of videogames entirely from scratch.

…this process is evolution, not learning. “Just like evolution in nature, the process isn’t really conscious of the overall direction it’s moving in. At each step, Angelina has a set of games to consider, and all it has to do is choose the best of this set, and combine their features together to make a new set,” Cook said in an email. Unlike, say, IBM’s Watson, every time Angelina wants to create a new game, the system has to start from scratch again.

Now it got me thinking about ACM and/or DCM (whichever term you prefer to throw around). If we dig around a few posts from the experts, folks like Max Pucher state that ACM is “not just runtime dynamic changes, but Just-In-Time creation of the process and resources WITH embedded learning, which means that knowledge of a previous case can be automatically used by people in a later case or process.” In further posts he links Complex Business Events to process and ACM which go into more detail which I won’t elaborate on here.

So what does this have to do with the games industry ? Well, simply put, I doubt anywhere in the BPM world are there people who spend all their time creating elaborate routines that display artificial intelligence in the way they do for gaming. Every large scale and major release focuses on creating more adaptive NPCs (Non-Playable Characters) and enemies, mimicking real life as closely as possible and challenging the player.

So, the question is: why can’t we hire these guys and get them plugging away on Process AI and our BPMS ?

The problem I have with ‘business rules’ being an inherent part of some BPMS is that they start to constrain process decision making and flexibility, but having a set of AI routines that not only adapts on the fly (or JIT) but also learn and potentially create a brand new process instance out of evolving from previous instances in the same or a variant way Angelina does surely will enable a more dynamic enterprise model to emerge too. Using Process Mining techniques to harvest the information that IT holds dear about where the actual process execution lies within could really be harnessed.

Could it be possible that this intelligent BPMS (God I sound like Gartner, shoot me) is far more emergent and where we need to be than just a set of analytical screens and real-time dashboards that they suggest ? Where’s the (r)evolution in that ? It still requires a person to conduct the analysis itself, make the changes. What if the system itself started to make those tweaks, or even suggest incremental or wide-scale changes ?

Make the intelligent choice and start headhunting.

Footnote: the picture is the Ferranti Mark 1 machine from the University of Manchester. In 1951 Christopher Strachey wrote a checkers program and Dietrich Prinz wrote one for chess. Arthur Samuel’s checkers program, developed in the middle 50s and early 60s, eventually achieved sufficient skill to challenge a respectable amateur. This was the recorded first instance of game AI.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s