Monday, November 25, 2013

Robotics Scientists Meet to Discuss Ethics

In October a group of leading robotics scientists and engineers met to discuss the ethical implications of their work.  Until recently a vast majority of industrial robots have been confined to cages where they work and humans are unwelcome.  Increasingly, however, we find robotic applications in our human world.  They are "aware" of us, sensing our presence, our identities, our activity and inferring from those data our intentions and goals.  And these robots act in our world too, based on their own objectives and what they can sense from their surroundings, including present human beings.


Now that they are among us, we need to ask some important questions:  What do we want our robotic progeny to do for us?  And where do we draw the line?  What do we want them to "know" about their environment in general but about us in particular, and, once they know it, what should they do with the data?  Who is creating these actors in our world and what what are THEIR objectives and priorities?  And what guidelines SHOULD they be following?

As usual, like an adolescent, our ability to act vastly outstrips our ability to understand the consequences of our actions.  Our ability to solve mechanical problems simply overwhelms our capacity to understand what we are doing to our world.

I'm not suggesting, however, that we stop acting while we somehow figure it all out in theory.  Also like an adolescent, we need to act to learn.  Experimentation is needed to acquire experience and real data.  We just want to be relatively certain that we know when we're running an experiment with potentially lasting or fatal consequences...

As I watch this happening from within the emergent, so-called Internet of Things, I'm somewhat ambivalent but very, very curious, perhaps cautiously optimistic.  I think we do need to proceed with the research for sure, but I'm positive we need to watch this carefully and make it visible.  I believe that regulations and testing are going to be important, as they are in other industries from mining and agriculture to aerospace and automotive manufacturing.  And I'm not sure that trying to slow progress or restrict research will have the desired results.  In other words, within the field, perhaps, we are doing all the right things in the domain of engineering.  But are we doing enough reflection?  As we drive science and engineering to achieve results, are we moving along as fast as we can in the parallel field of ethics?

What we need in my opinion is a significant investment in the field of philosophy -- ethics in particular --- from research to public outreach and engagement with scientists and engineers alike.  This is not unlike the importance of ethics in biology with respect to genetic engineering and medical technology.  We shouldn't expect meetings like this to definitively answer such profound and difficult ethical questions.  But on the other hand, we should celebrate their efforts, encourage more meetings like this with different populations and diverse points of view, and insist on transparency and the opportunity to inform ourselves.

Science writer John Markoff got it right when he observed that although the conference did not produce any clear answers, it seems useful "to persuade the researchers to confront the implications of their work."  I only wish the article had gone a little farther in this dimension and given us a feel for what kinds of implications they considered.

No comments:

Post a Comment