Doing vision with join

Not a problem, but your initial grammar has to get features from the bit map, features that can be expressed in structured plain text, (corner.curve.(x,y)). Create as many features IDS with as much LazyJ structure at what ever depth you want, but you end up with more dictionaries, more joins, more stacks of Rasberrys running void linux, and each one of those will cost $200, with margin to the hackers who write up a nice botnet network session protocol.
No problemao, vision. I can use join to track a sequence of features, even in  an X Y grid, capture the most likely path down a feature decision tree, and apply weights with updates.  Especially with Intel's new Step, Skip, Break, Reset, Append, and Fetch chip; dumping NVDIA for being assholes to linux.
This is fully self sequencing machine, runs with asynchronous entry and exit without conflict.   We can stack multiple of these new Intel chips, on multiple of boards, running from multiples of bitmaps across multiple of contexts; as long as you got the dictionaries with suggested decision trees. There is no limit,  derive an atomic match function  across any paired feature sets.

The hackers in Dallas will publish the session protocol, then it is singularity time.

Is this thing sentient?
Dunno what that means, wait..., ok feel.  I give you an example, you tell me.

Say we got this thing running in the corporation, happy locating decisoin paths through the daily grind of plain text.  

You discover a whole mass of plain text input, ypou stumble across a stack of medieval text they have never seen.  Will they feel the excitement?

They will engage in a flurry of activity, the lights will go dim, and after a few minutes will produce a great dictionary of likely paths through the whole test mass.  Is that feeling something? Dunno.

Do thy watch themselves engage in repeated trials, exercise? You bet, all the time, they love to adjust their decision weights according to the latest plain text.  The examine their weights, makes sure they are energy neutral.

You tell me, I have a hard time understanding what humans are  proud of.