Tag Archives: artificial intelligence

Computer Technology: Art or Science

Over the last week or so I have been troubleshooting some problems within my home network.  As a result I have come to wonder how much managing computers software and hardware is an art versus a science

Tools and Usage

I have come to realize the tools themselves are exact and follow  a systematic methodology.  In that sense they are exact, for example, either on or off, do this or do that.  

I found though how you use them has many variations and subtle nuances where emotions (the artsy side) can play a role as much as logic (the science side).

Background

My web sites have be set up to use encryption via secure socket layer (SSL) with authentication provided by an independent Certificate Authority.

Basically content exchanges between someones computer and mine (both to and from) is encrypted so it can only be read by the intended recipient.

Also through the use of a certificate granted by a trusted source called a certificate authority (CA)  the sender can be sure the computer they are connecting to is mine. The CA is Let’s Encrypt in my case.

If the certificate is not valid your browser will warn you about the potential perils of proceeding – i.e, the target computer might not belong to who you think it does.

The certificate authority has a process I must follow every 90 days  so they can be confident I own the domain name and associated compute.

What Happened

Although my certificates have been in place for over 4 years (renewing every 90 days as required) something went wrong recently and my websites and servers were shown as unsafe.  (That is I had no valid certificate).

Without going into great detail, which would bore the majority of my audience, the point of the story is my troubleshooting which lasted over a week was fraught with many illogical, but perhaps artful steps.

In the end, I did it.  I finally got a solution. What interested me most though was the fact I needed to really dance around to get what I wanted (A valid certificate installed on my server).

It turns out this was not just a do this/do that type of recipe with straightforward repeatable instructions, which technically is what you would expect.

Artificial Intelligense

This experience has given me a better understanding of some of the pitfalls of artificial intelligence.   As computers think more like humans the art of behavior (emotional motivation) versus the scientific ways of acting will come out more and more.

What language do you think in

Thinking is an individual process centric to the human brain.  Basically the construction or development of a particular opinion, belief, or idea about someone or something constitutes thinking.

Artificial intelligence represents the ability to think without a human brain.  As the description of thinking is very broad, measuring if it is occurring can be a very subjective practice.

Alan Turing, British computer scientist, academic (among other things) postulated a test (the turing test) for artificial intelligence . It is however a somewhat subjective test.

Do you think in language?

Thinking is more a concept, that is thoughts are conceptual (e.g. think of sound or smell).  Language is a communication method associated with the thoughts but some  cognitive experts argue language is used more as an internal monologue tool than as a requirement for thought.

I have a number of multi-lingual friends, colleagues and family within my personal network and the simple question, “What language do you think in?” often leads to puzzlement.

Most, in fact initially everyone I asked was not sure what the question meant, hence the puzzlement.  On the whole, my conclusion is language is not required for thinking.

That being said, the ability to communicate is a key requirement for assessment  that artificial intelligence (thinking without an organic brain) exists and that thinking is occurring.

Still I wonder?

  • Can thinking exist without language, even if that is sign language
  • does a computer or device need to be able to communicate to prove the presence of intelligence
  • does thinking require the ability to comprehend and is language required to do this
  • if  I am multi-lingual and having a internal monologue, which language do I use.

And the questions continue.

 

 

The Robots Among Us

Several years ago Google developed humanoid robots with embedded artificial intelligence.  The exterior skin and mechanical movements were so realistic they appeared fully human.

Google has kept extremely tight security on this project and have been able to successfully keep it secret from the general populous. This has enabled the to run a pilot without interference.

Google Support and Sales Reps

Gradually google introduced their andriods into the Google work force hiring them as sales and tech support representatives.  Only a small limited cadre of Google insiders know of this pilots existence.

Amazingly, none of the robots have been identified as such.  So far everyone they interact with, whether inside Google or external is unaware they are actually working with an artificial being.

The Resistance

A small group of humans have become aware of Googles experment but remain underground out of the public view out of fear for there own saftey.  Google is not ready for the world to know how technologically advanced they have begun.

The small group that is aware of the robots among us have developed special tools, worn as wrist watches that detect the prescence of the artificial beings.

This helps make sure those in the know don’t share inappropriate information giving away the fact they are aware of the robots among us, putting themselves in danger.

Fantasy

Is this post real or fantasy?  Hmmm.

Maybe you should test that next person you meet with probing questions, study their emotions.  Can you tell if they are born or built?

Artificial Intelligence

Saw a great movie a couple of weeks ago called ExMachina.  It was an exploration of the concept of artificial intelligence.  That is intelligence made or produced by human beings rather that produced naturally.

Allan Turing was a British scientist considered by many to be one of father’s of computer Science.  Among a number of significant achievements before committing suicide Turing  introduced what is now known as the Turing Test in a paper he wrote that was published in 1950.

In essence he proposed that if a human examiner was set up to converse with other humans and machines (through the same interface) when the conversation with the machine could not be identified as such, the machine could be deemed intelligent.

In ExMachina the story was a variation on the Turing Test.  The tester had only one subject to examine, an android.  In a twist the builder of the android designed the test to see if the machine could manipulate the tester.

In 2014 PrincetonAI’s submission to the annual Turing Test (a machine called Eugene Goostman) fooled 33% of the judges.  There is of course controversy associated with this result – did the developers games the system?, Was the test too easy?

Regardless of the questions, Artificial Intelligence is clearly on it’s way.