AI and Ethics | Part One

What is the future for Artificial Intelligence and what does it mean for our ethical and legal approach to it? [Part 1]

This was an essay for the Media Technologies module at SHU.

Do you think that we’re close to co-existing with robots or is that something for your grand-grand-grandkids to worry about? Technology’s advancement has had an incredible development in the recent years. When looking at the patterns of progress in the past, we can see that over time, incomprehensible society advancements are increasing. The progress in 2016 is a much bigger incomprehensibility shock from the technology level in 1750 than ever before. It has been brought to attention that a person from the 1750’s visiting 2016 would struggle a lot more to comprehend the developments and society, than someone from 1500’s visiting the 1750’s – despite the two timeframes being very similar. (WhyFuture, 2017)

smart-home-3148026

What is believed to be the most important technology of this era is Artificial Intelligence, otherwise referred to as AI ( Brynjolfsson and Mcafee, 2017). It is a system that can be built to be able to process and respond to data in ways that are seen as human-like (Lacoma, 2017)

With the definition vaguely broad, it includes a large variety of different types of AI’s that are currently being used, developed or still in our future imaginations. and can be classified in different ways, such as General and Narrow(Lacoma, 2017) – where it either mimics human behaviour, for example Sophia the robot (Sharma, 2017) or is just focused on a specific problem, area or action, such as self-driving car. Throughout these blog posts, I will discuss the future of Artificial Intelligence and how that affects potential ethical approaches that need to be considered.

When it comes to fast technological development in AI, it’s important to refer to the theory of singularity. It is a prediction of a point when their intelligence will match the human mind abilities and then go above them reaching beyond our intelligence (Ghose, 2013) as it would have the ability to ‘upload’ more intelligence to its mind and redesign and expand itself. Ray Kurzweil believes that it will exceed the capabilities of human mind, even the areas where we are still superior to them – as they are already capable of going beyond our abilities when it comes to find cures or solutions to economical or environmental problems (IEEE Spectrum, 2017) Once AI gets on the step above humans on the intelligence ladder, it is believed that it will then be able to self-develop its own abilities further, and each time it takes another step, it will reach even further, more powerful and faster abilities to self-improve, thus creating a Super-Intelligence. (Hunt, 2016). In support of this theory, Solomonoff (1985) explains that AI machines would eventually reach an ‘infinity point’, where ther capabilities would increase infinitely.

Although the majority of experts believe that we are still a long time away from the moment that AI surpasses the abilities of the human mind (Brown, 2016), many leading academics and research centres predict the development of fully conscious machines to be as early as 2030, and state that preparations must be put in place as early as possible, as by 2045 it will be billions of times more powerful and intelligent than humans (Lu, 2016; Ghose, 2013). However, the idea that powerful AI will develop in the near future, which has even been supported by minds like Stephen Hawking, is still facing a lot of criticism. Earnest Davis, who’s a New York University computer scientist, sees no signs of being close to singularity and is definite that current AI is light years away from even reaching a child’s level of human-like mind and consciousness (Ghose, 2013).crew-22235.jpg

Similarly, Brown (2016) also states that it is unlikely we will have a general intelligence of such measure anytime soon, due to the complexity it will have to reach, in order to match the human brain’s history of development spreading over millions of years.

However, these points seem to only take into account what has been achieved so far, as comparison and a basis to the level that needs to be reached – as in, so far it has taken us this long to create what we have, meaning we’ll need way much longer to achieve the stages we’re talking about, as they are on such a high level. Looking at it like this, they seem to lack consideration for not only the speed of technological development that is continuously increasing, as the same amount of progress that was made before, will now take less and less time. They also seem to overlook the fact that patterns of technological developments don’t match the patterns of human development and evolution. Therefore, we can’t predict the developments of artificial intelligence accurately by comparing it to how human intelligence was formed – and neither can we predict it based on its previous progress.

For these experts being, just like the majority of people, linear thinkers, it’s what makes makes them believe that AI singularity is as far as Centuries away. Instead, it’s important that we see how the rapid technology development, which makes human evolution very slow in comparison, does not progress in a linear fashion. This can be seen through the patterns in progress that were previously pointed out when comparing the differences between the progress in the time during 1500-1750 and 1500-2016. It is clear that incomprehensible shock difference is dropping over time, as the progressed between 2000-2014 was equivalent to the progress of the entire 20th Century. Based on that, it can be predicted that that same 20th Century amount of development will eventually occur many times per year, then months and then even days (WhyFuture, 2017).

Hunt (2017) believes that, because of this, what now seems impossible will eventually become possible, if not easily achievable. He proves the failure of linear thinking by pointing out the lack of alignment between the predictions of film and tv with reality, such as Star Trek not using touchpad interfaces until the 24th Century.

This point could, however, be challenged by the fact that some science fiction actually predicted much better technology improvements than achieved in reality, such as the world of Blade Runner (1982) set in 2019, which is now only a year away, and yet we are nowhere near the technology portrayed in there, which could suggest that many experts may still be overestimating the future of our technological advances.

markus-spiske-207946

However, real examples of AI can be used to prove that our progress is speeding up way beyond our expectations. AlphaGO, which is an AI machine used for playing the game Go, impressed the AI community with its win, as academics believed that it wouldn’t be cracked for at least 10-15 years (WhyFuture, 2017). A previous impressive AI developed by IBM for playing chess, Deep Blue, which achieved a shocking win when defeating Garry Kasparov in 1997, would have taken longer than the ‘entirety of the universe’s existence to calculate GO using its own method (Hunt, 2017). Therefore, AI’s progress has already proved to speed up past everyone’s expectations. By using exponential thinking, rather than linear, to adapt to the current progress patterns, we can see that singularity and super-intelligence may indeed happen in the near future. Some scientists believe we are currently already in the midst of the singularity, as even things that can be considered the core of humanity’s values, whether it’s creativity, caring for the elderly or sexual pleasure, have been outsourced to machines. (Ghose, 2013)

When you realise just how close we are to incredible Artificial Intelligence inventions and the potential of their future abilities, many ethical issues come to mind. In the next blog post, I will consider how ethical approaches link with the current and future rise of intelligent machines.

Read Part Two here.


Bibliography:

Anderson, M. & Anderson S.L. (2011). Machine Ethics.

Bidshahri, R. (2016) If Machines can think, do they deserve rights. Singularity Hub. [Online Article] Retrieved from https://singularityhub.com/2016/09/09/if-machines-can-think-do-they-deserve-civil-rights/#sm.0001u3nvkil9id6atzl1x5c1vmdj9

Blade Runner (1982). Michael Deeley (Producer). Ridley Scott (Director). Warner Brothers. [DVD]

Bossmann, J. (2016) Top 10 ethical issues in Artificial Intelligence. World Economic Forum. [Online Article] Retrieved from https://www.weforum.org/agenda/2016/10/top-10-ethical-issues-in-artificial-intelligence/

Bostrom, N., & Yudkowsky, E. (2011). The Ethics of Artificial Intelligence. [Draft for Cambridge Handbook of Artificial Intelligence, eds. William Ramsey and Keith Frankish (Cambridge University Press, 2011)] Retrieved from https://nickbostrom.com/ethics/artificial-intelligence.pdf

Brynjolfsson, E., & Mcafee, A. (2017) The Business of Artificial Intelligence. Harvard Business Review. Retrieved from https://hbr.org/cover-story/2017/07/the-business-of-artificial-intelligence

Dormehl, L (2017) I, Alexa: Should we give artificial intelligence human rights. Digital Trends. [Online Article] Retrieved from https://www.digitaltrends.com/cool-tech/ai-personhood-ethics-questions/

Ghose, T. (2013). Intelligent Robots Will Overtake Humans by 2100, Experts Say. LiveScience. [Online Article] Retrieved from https://www.livescience.com/29379-intelligent-robots-will-overtake-humans.html

Hunt, D.G. (2016) The future of artificial intelligence and ethics on the road to superintelligence. [Online Article] Retrieved from http://www.whyfuture.com/single-post/2016/07/01/The-future-of-Artificial-Intelligence-Ethics-on-the-Road-to-Superintelligence

IEEE Spectrum (2017). Human-level AI is around the corner – or hundreds of years away. [Online Article] Retrieved from https://spectrum.ieee.org/computing/software/humanlevel-ai-is-right-around-the-corner-or-hundreds-of-years-away

Lacoma, T. (2017). Demystifying artificial intelligence: Here’s everything you need to know about AI. Digital Trends. [Online Article] Retrieved from https://www.digitaltrends.com/cool-tech/what-is-artificial-intelligence-ai/

Lu, C (2016) Why we are still light years away from full artificial intelligence. TechCrunch. [Blog Post] Retrieved from https://techcrunch.com/2016/12/14/why-we-are-still-light-years-away-from-full-artificial-intelligence/

Sharma, K (2017). We’re all getting played by Sophia the robot. Fortune. [Online Article] Retrieved from http://fortune.com/2017/10/27/sophia-the-robot-artificial-intelligence/

Solomonoff, R.J. (1985) The Time Scale of Artificial Intelligence. Human Systems Management. Vol. 5

WhyFuture (2017). The Future of artificial intelligence and ethics on the road to superintelligence. Retrieved from https://www.youtube.com/watch?v=8Ja_7Fx2MmU

One thought on “AI and Ethics | Part One

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s