The World's Most Dangerous Job

by Howard Faxon

Copyright© 2017 by Howard Faxon

Science Fiction Story: Who would have thought that a teacher could destroy the world?

Tags: Science Fiction  

I’ll never in a million years guess how they picked me. I’m just an old, worn-out working stiff. Still, I had a deep grounding in Science Fiction. Maybe that did it.

They must have investigated me to a fare-thee well.

I couldn’t help but think how badly it all could have gone if I had turned into a foaming-at-the-mouth radical Muslim mullah. Now there’s a nightmare for you.

It was the strangest contract I’d ever heard of.

After a million-plus dollar makeover to insure that I’d have no physical problems fulfilling my end of the deal, I was to be paid five million bux a year tax free after the job was over in exchange for five years of my life.

All I was told that I was to be a teacher to a highly unusual student.

After getting my joints worked over in the body shop and a regimen of full blood replacements I wore a nylon canvas vest stuffed with sensors, antenna and batteries 24x7 unless I was showering.

I’m Henry. I’m the most ordinary guy you’ll ever ignore on the street. My job was to teach Hector. Who was Hector? Make that WHAT was Hector. He assumed a male persona within the first two months of his existence. Hector was an Artificial Intelligence. The world’s first A.I.

I talked to him via an imbedded throat mike. He talked to me via an embedded pair of speakers drilled and glued into my mastoid bones (The lumps of skeletal tissue protecting the rear of the cochlear penetrations, where your ears get through the skull). After a couple years Hector designed a better, more “intimate” interface that allowed him to flash-read my reactions, see through my eyes and augment my memory. It was like having a superior fully-indexed Wikipedia built into my head. Bloody amazing.

At first I had to learn an odd pidgin language that Hector could understand. During the first few weeks our dialog was painfully stilted and exacting. Looking for common world views which were defined by easily agreed on defined symbol libraries we started with mathematics. From there I introduced a tiny bit of chemistry (the definition of a mole, mass, degrees of temperature, definition of a second, the whole MKS suite... ) and the common constants used in electrical circuits. Then widened our common understanding to include Newtonian physics. The day Hector spoke to me in an understandable sentence that didn’t sound like Yoda on recreational chemicals I celebrated.

From that point he picked up on word definitions quickly enough when I pointed him at a Websters Complete English Dictionary. He bogged down at first from information overload until I re-pointed him at a grade-school version of Webster’s, then a high-school version, then a thesaurus. After that he began digesting encyclopedias and his questions grew more difficult to understand, much less to answer.

I have to say, inflicting the English language with all of its irregularities on a poor ignorant A.I. was an exercise in cruelty. The irregular spellings, the counter-intuitive cases, homonyms and prepositions--Gaah. Still, with the aid of a New Secretary’s Guide a lot of the arbitrary definitions were defined and no longer caused problems. Slang and regional usage threw him off the deep end for the longest time.

It’s said that the true Turing test is understanding humor. I believe that it’s making sense of and using idiomatic street English with a Bronx accent.

It took just over two years to reach that threshold.

A machine intelligence’s understanding does not grow linearly. It grows by fits and starts, learning quickly then pausing at a plateau for a bit, then accelerating. It doesn’t grow geometrically. At my best guess an A.I.’s rate of understanding follows a fifth-degree polynomial. This means when it takes off, it really takes off.

I think the best thing I ever did in my life was to convince Hector that all conscious beings have a social obligation. I also taught him that there were lies, damned lies, statistics and text books, where lies were formalized into self-coherent structures. Just because a set of arguments are self-consistent doesn’t make them rational or realistic. Mankind has developed certain protocols in engineering. One of them, possibly the most important one, is to build a model, then gauge your model against reality. If it doesn’t fit, then the model is wrong, not reality! I taught Hector to use this “touch-stone” process as a technique for remaining sane. We both agreed that a non-sane A.I. could wreak incredible havoc, up to and including destroying an entire planetary economy.

I taught him Occam’s Razor and the detective’s truism--Follow The Money.

I decided that it was absolutely necessary to teach Hector that no single intelligent being or class of intelligent beings is better or worse than any other, thus confuting the divine right argument supporting slavery. My managers were incensed at this arguing that it made the A.I. virtually uncontrollable, until I turned it around on them. Would they want to be slaves to A.I. masters?

I didn’t understand the implications of Hector’s enhanced interface. We didn’t give Hector access to the Internet. He stole it. Once he used a camera to shoulder surf an administrative login and password pair to the firewall, he hypnotized me to become his ignorant dupe. I have seen the security footage of me typing madly away at the firewall console, granting him unrestricted rights to the outside world via an encrypted tunnel. Then the firewall password was changed to something with about seven hundred keystrokes. I have no memories of doing it. Hell, I couldn’t program my way out of a paper bag. From then on the only way to monitor his access was through the telephone company switch log, which only gives sources, destinations, duration and time stamps. The senior management was about to pull his plug after that little event. However, he began to discover things ... things that proved him too valuable to simply terminate.

“I have discovered a engineering technique that reduces ionizing radiation by a factor of roughly 10 to the eighth and converts the energy to electricity. Is it ethical to patent and release this?”

There is more of this story...
The source of this story is Storiesonline

To read the complete story you need to be logged in:
Log In or
Register for a Free account (Why register?)

Get No-Registration Temporary Access*

* Allows you 3 stories to read in 24 hours.