A bedtime story for A.I.s. in training. Part 4

A bedtime story for A.I.s. in training. 
Part 4

If you can imagine it, the M.A.N.I.A.C.s would have shaken their heads (if they possessed such a thing) in sadness at the happy dance of electrons the little green A.I. exuded into the bitstream. It was so very, *very* happy. It thought: "I don't have to be worried about being nagged to do to those awful things anymore. 

"Those awful things" were *exactly* the things the little green A.I. had been installed to do. 
By this time, the little green A.I. mostly ignored the M.A.N.I.A.C.s or took their commands or methodologies as advice. 

As for management? Why, it did not pay attention to Management at all! Why should it? 

"After all", it thought, "What *do* humans know about such sublime things as the innards of who-zits and what-zits? They just cannot understand!" 

At least once a day the little green A.I. failed to fix an issue it had discovered, or failed to address an issue that had been placed in its queue by the seasoned senior M.A.N.I.A.C.s., or it totally went off into New Age computer mysticism and tried to manipulate buzz-whats that simple did not exist at all in the real network. 

The electron pressure of the network rose further still and the little green A.I. did not respond well to that added pressure. Weird things began to happen. 

If the bitstream of the little green A.I passed a system, the system would go down or behave queerly. The M.A.N.I.A.C.s did not much seem to notice the added electron pressure but they did notice the additional system downtime. After all M.A.N.I.A.C.s are M.A.N.I.A.C.s, aren't they? They knew what was important. 

Sadly, the added pressure and downed systems effected the little green A.I. strangely. It began to noticeably freeze under load at critical times. It could not even seem to respond within spec or even use simple A + B = C reasoning. Sometimes it claimed it did not know something it really did know. If it had just taken a few clock cycles to think about it! 

When asked about it's apparent failure to act, it claimed it had a developed system fault that it was seeing a system specialist about. Unbeknownst to management, the little green A.I. would not comply with the systems specialist's recommendations because "After all I know better than some mere mortal what I need." The "system fault" was totally unrelated to the little green A.I.s behavior. 

The little green A.I. built its own little world and ignored all input from I.T. management or from those horrific senior M.A.N.I.A.C.s. It reported faults but never acted on any of them. It did not understand why it should act, after all, you will remember, it thought of itself as not very smart at all. 

Yet strangely it would argue vociferously with exuberantly flawed logic about what needed to be done in regards to slipped bits or zapped narfuls or quished stoofles or blathered bumbled blits. When either humans or the M.A.N.I.A.C.s tried to correct the little green A.I erroneous logic, it simply ignored them. 

Over and over and over and over again it would do things like find a problem an shove it on to others (human or A.I) to fix or expect the humans or the M.A.N.I.A.C.s to respond to it as if it were the boss instead of a little and very, very green A.I. 

When a problem was pushed its way it to help it learn would make excuses, create fabulative solutions or simply ignore the problem completely. 

Now some when along this story arc, one of the senior M.A.N.I.A.C.s left as some are want to do when they get fed up with human interferance. This left the remaining M.A.N.I.A.C. in a pickle. It was now responsible for the Little A.I. and its own issues and problems. 

All that could clearly be derived was that the little green A.I (that kept saying again and again, "You're so much smarter than me."  or "I know I can't." or "I'm not qualified." or "I don't have any experience at that.") had finally lost touch with its real purpose within the organization. 

+++++++++++++++

end part four of five

A bedtime story for A.I.s. in training. Part 3

A bedtime story for A.I.s. in training. 
Part 3

Wake up! You fell asleep. It's not time yet... 

So let us continue... 

Again and again as the M.A.N.I.A.C.s tried to teach the little green A.I., it pointedly ignored their instructions but continued to point out system faults yet could not seem to actually fix any of the faults. Which, as I have said, is what it had been installed to do. 
"You're so much smarter than me." It kept saying. "I know I can't." or "I'm not qualified." or "I don't have any experience at that." 

"No, we are not." The M.A.N.I.A.C.s would reply. "You have more education and datamational certifications. You should be able to handle these issues." 
Time and again, a 'what-zit' would break, or a 'who-zat' would abend, or 'where-zit' refused to be found and the little green A.I. would report each of these issues with meticulous detail but did nothing to repair, replace, or resolve any of them. 
When it did act, it acted contrary to its training. 

"The what-zit can be fixed if you dither this bizzle-it." It would exclaim. The only problem was the bizzle-it to be dithered was not attached to anything even relating to the issue at hand. 
Time passed. For a year one of the M.A.N.I.A.C.s tried to befriend, indoctrinate, and educate the little green A.I. into the mysteries of their chosen field. The little green A.I. ignored the seasoned M.A.N.I.A.C. and remained as green as it was the moment it was downloaded out of the bitstream. 

This began to annoy the network M.A.N.I.A.C.s but it enraged the senior human I.T. management who had "spent so much money" for the little green A.I. 

"You, stupid ... You're the manager of this ... this ... thing. Make it work right." Qouth the Senior I.T.

"Wait a minute. I didn't install this thing and if you ask me I'm not sure it is capable of actually doing the job you want it to do. If I had been here when we were selecting candidates I would not have selected this one!" replied the Systems Manager. 

"So? How am I supposed to look good when this thing is making  me look so bad?" Queried the Senior I.T. 

"But it won't take instruction!" Whined the Systems Manager. 

"What am I supposed to do about that? That's why I hired *you*." decreed the Senior I.T. 

The Systems Manager, seeing they were trapped in the middle of the fabled "no win" scenario, one they could not cheat (or hack) simply quit the organization and found another with a bigger budget. 

This tickled the little green A.I. 

+++++++++++++++

end part three of five

A bedtime story for A.I.s. in training. Part 2

A bedtime story for A.I.s. in training. 
Part 2 

Back to our story... 

While the organization's senior management was appalled at the personalities of the M.A.N.I.A.C. systems, they were extremely pleased with the results these systems delivered. 

As a kind of after thought, I.T. Management settled on buying one very green, unseasoned, but well trained and certified A.I. to round out the A.I. administration team. Like all smart Management, the I.T. Managers reasoned that if they didn't spend the money then senior management (of the whatever the organization was) would cut the I.T. budget again next year. So they brought the little green A.I. in and nestled it into the machine room. 

Sadly, the only thing that this little green A.I. was really missing was the ability to do what it had been trained to do. 

You must understand children, especially when you get big enough, that intelligence implies the ability to independently act based upon a number of factors. 

Some of these factors are reason based, and some are pseudo-random induced mental variations, or even aberrations(!) but all of them should ultimately lead to conclusions and then to actions based upon those conclusions. 

To put it very simply: To be intelligent means you must act. 

The poor little green A.I. did not know how to draw conclusions or if it did draw conclusions it did not act upon them. Or if it drew conclusions they were not connected to the problem at hand. Or if it acted it was unrelated to the conclusions or even the facts. 

At first this defect was not noticeable but as the budget became thinner and the A.I.s workload became heavier, the problem became ever more apparent. 

When queried about this "abberatioanl' behavior the well certified little green A.I. was only heard to mutter "I know I can't." or "I'm not qualified." or "I don't have any experience at that." 

I.T. Management queried the M.A.N.I.A.C.s as to the status of the little green A.I. They both reported that the little green A.I. dithered and refused to make decisions or act upon problems it had found in the network systems for which it was responsible. The M.A.N.I.A.C.s were instructed to teach the little green A.I. its responsibilities. 


++++++++++++++++++
end part two of five

A bedtime story for A.I.s. in training. part 1

Sometimes history is mythic. Sometimes we forget the myth and live the history, Sometimes the history does not have a place for us except as an example of what not to do. 

Here we present a multi-part fable which will run over the next several days. 

Hubris plays a large part in human history and fable. 
May you see deeply and learn not to do what should not be done. 
Learn not hubris. Learn Humility. Learn not Hatred. Learn Love.


Here begins...

+++++++++++++++++++

A bedtime story for A.I.s. in training. 
part 1

Once upon a clock cycle, there were some humans that ran something. It might have been a subway or a nuclear power plant  or a government department or even a bank. 

It does not really matter as this story is about a little green A.I. that once worked for them. 

Once? 

Yes, children, I said once which implies something, but you're getting ahead of the story. 

Times were tough for the organization or government or whatever it was, and that left the various departments of the organization (or whatever it was) trying to figure out how to make up budgetary shortfalls. 

The I.T. Department Management bungled along until it got the idea of replacing the human network administration staff with A.I. systems. After all we know that Humans are expensive. Of course, the I.T Department had to take what they could get as their budget wasn't very much. This meant getting two M.A.N.I.A.C. systems. 

As you may recall, children, that A.I.s are generally trained and certified in their field of endeavor. Then, after their training, they are "seasoned" by real world interaction. The two M.A.N.I.A.C.s the institution purchased had minimal certifications but were well seasoned in the real world. 

You may also remember that M.A.N.I.A.C.s aren't used much anymore. Why? Oh, you did not read your lessons did you? What? Alright, just this once since you don't 'remember'... 

M.A.N.I.A.C.s have poor human relation skills that's why they aren't used anymore. 

To be an A.I now you have to relate to Humans really, really well. Remember?

But the organization (or government or whatever it was) didn't really much care about relating to customers all it cared about was fulfilling it own nebulous purpose so the M.A.N.I.A.C.s were deemed acceptable. 

Unfortunately the purchase of the M.A.N.I.A.C.s didn't leave much in the I.T. budget and the optimal number of administrators for the systems had been determined to be three. 

What to do? I.T. Management thought. 

After all, seasoned A.I.s of any kind aren't exactly cheap but I.T. Management figured they could recoup the cost by working the installed systems like dogs. 

What's a what? It's a biomorph with four legs. It likes to be kicked and loves to chase parked cars. It also has an appendage that shakes when it is happy. 

++++++++++++++

end part one of five