Back to the Future

OK Fans… My latest Display Daily editorial, the one on uLED based displays is now available:

http://www.displaydaily.com/display-daily/29427-back-to-the-future

Is Artificial Intelligence a Threat to Humanity?

Recently, pundits and scientists alike have been wringing their hands about the “threat” posed by Artificial Intelligence, even though we seem to be no where near being able to build anything remotely like a super-intelligence.

http://thebulletin.org/artificial-intelligence-really-existential-threat-humanity8577

What I find fascinating is not what is being said by whom, but that those who are talking about this seemingly fail to take note of the fact that Science Fiction has been exploring this issue far longer and deeper than the technologists of today.  Somehow, its as if this topic was totally new, never before discussed.  I find this especially true in the area of ethics.  Seriously, why do we not see any discussion about such concepts as Asimov’s Three Laws?

Also, why is this discussion so pessimistic?  Why is that we don’t see the possibilities that AI will be a boon?  Instead, we see articles that the machines will “take over”, they will make us obsolete, they will self-perpetuate and be selfish, etc.  Why should that be the case?  Have they never read Asimov?  No, they have only seen that rip-off, twisted Hollywood version of I Robot.

Or they react with horror at the thought of super robots and drones in the battlefields of the future… and yet, not one of these pundits seems to have ever read Keith Laumer’s Bolo series?  Never read, Honor of the Regiment?  They haven’t read of the story of the first fully self-aware BOLO that saved a world by refusing a technically lawful, but unethical, order knowing that that refusal would trigger a computer worm that would eventually, in an hour or so, “kill” it?

I see a vastly different future in which AI, based on biomicry of the human (and other species) brain will be partners with us, capable of doing things that we can’t, like survive a thousand year journey to the stars to terraform promising worlds into new homes for humanity.  This is at the heart of my upcoming novel, All the Stars are Suns.

The Naked Brain

In the SciFi novel I’m writing, All the Stars are Suns, set into the future, neuroscience has advanced far enough that we can model and build biomimetic analogs of neural functions, to the point were we can fabricate inorganic artificial brains.  These brains, being biomimetic would even have human emotions, if we so chose.  Sounds too far fetched?  Obviously, I don’t think so, or I wouldn’t be including them in my story.  Although I believe that we are many decades, perhaps even centuries away from truly fabricating human like inorganic brains, the state of the art today is perhaps a lot more advanced than you might think.

For example, take a look a this video from NATURE:

I imagine that someday we will be able to convert images like this into wiring diagrams.  No, I’m NOT a transhumanist.  I do NOT believe that this will allow anyone to “upload” their personality and memories to an inorganic copy of themselves.  Instead, I foresee that through imaging many brains we will come to understand the basic functions of the neural nets and model them, even create physical instantiations of them, which will allow us to fabricate sophisticated neural net computers that function much as our brains do.  With experience, they will learn, just as humans learn.  They will be themselves, not carbon copies of us.  And they will have their own quirks, since they will not have had the ongoing learning experiences while their brains rewire themselves as ours do from infancy to adulthood.  They will be “born” already mature, though untutored.  I’m exploring the ramifications of that type of “growing up” in my novel.

(Addendum 10/8/2015:  Here’s a paper on the development of a computer simulation of a TINY portion of the neocortex of a young rat’s brain.  Note that they needed a supercomputer to run it.  Thank goodness for Moore’s law.  Maybe someday we will be able to run such simulations on computers available on a start-up company’s budget.  Someday, neuroengneering will be a ‘thing’:  http://www.eurekalert.org/pub_releases/2015-10/cp-rba100615.php )