What are the connections between ethics and artificial intelligence?
What are the issues that have arisen, especially in the business context? What are the issues that may arise?
Artificial Intelligence Stack Exchange is a question and answer site for people interested in conceptual questions about life and challenges in a world where "cognitive" functions can be mimicked in purely digital environment. It only takes a minute to sign up.Sign up to join this community
The connections between ethics and artificial intelligence can be divided into five major categories, and other categories may form over time.
Since the most important in the long term is the most likely to be dismissed by those with normal perspectives about ethics and AI, the four will be addressed in reverse order
Automatic Detection and Remedial Action
The pattern recognition capabilities of existing AI systems and sub-systems is already employed to detect a variety of ethical breaches.
Remedial actions may be the opening of a case with the automatic generation of a notice to those in potential breach.
Smart Organized Crime
Although much detail could be included here about detection avoidance in crime using AI, it may not be socially responsible to include such in a global public facing site.
Threats to Economies and Individuals
As with any high impact technology, disruption is a possibility. This was true of fire, irrigation, the wheel, bronze smelting, gun powder, typesetting, steel-working, engines, textile automation, alternating current power distribution, aeronautics, petroleum refining, electronics and radio transmission, pharmacology, terrestrial nuclear reaction, and the Internet. Genetic engineering and artificial intelligence are next in line.
What ethical conventions will likely be impacted?
All of these have either direct or direct impact on the viability of business options and how business may be conducted.
Some may consider dominion over the earth as an ethical grant to humanity. Others may equate the soul with the cognitive and self-aware aspects of only one species on earth.
We already have a world that is sufficiently disconnected on an existential plane that many consider their cats or dogs as more important than any human. When people are more connected to their intelligent agents than their pets, family, and friends, that may qualify as an ethical impact. Others may see it as a psychological impact.
Realistically, it is ontological. What is a human to think of her or his only purpose when it becomes questionable whether homo sapiens is simply a link between DNA based intelligence and some more capable species the reproduction of which has been decoupled from DNA coding.
Replacement of jobs have caused changes in what families wish for their children. What will be the impact when few job roles (or eventually none) exist where artificial employees don't exceed their human counterparts in effectiveness?
If humans cannot adjust to the idea that the sole purpose of life has nothing to do with practical provision of water, food, shelter, clothing, and essential products and services, there may be systemic depression. Conversely, leisure may become the reality for all humans, leaving ethics and the finite nature of DNA based life the only two concerns of humans.
Correlations Between Ethics and AI
This is the most unpalatable of categories to examine when the examiner is human. It is possible that artificiality may be an ethics progenitor. The limitation of humans as ethical beings is well documented. It is possible that AI may be more ethical than its developers.
Will a group of AI systems be able to arrive at method for distribution of power and a standard for global trade that is as good as or better than humans have been able to negotiate and then police each other in a way that leaves no possibility of undetected breach of treaty?
This is a good related read from Nature: There is a Blind Spot in AI Research
Fears about the future impacts of Artificial Intelligence are distracting researchers from the real risks of deployed systems
In a business context, there are issues surrounding the implementation, the implementors, the other employees, the business entity itself, and the customers. These stem from data used, risks inherent in an implementation, like unknown errors bugs or algorithms without human checks, behaviour change of impacted stakeholders, job losses, reputation impacts on the company, etc.
There's a lot to think about AI and ethics. It could be seen as a combination of a broad set of topics including computer science, humanities, economics, and philosophy.
I run a podcast on some of these issues http://machine-ethics.net/ let me know if there's anything you want to be discussed or someone you would like to hear.
With computing in general, optimization is a major goal. For AI, optimal decision-making is what it's all about. But sans humanity, this may prove to be problematic.
(Apologies for the brevity--I'll be returning to elaborate since this is a subject of personal preoccupation--but I wanted to leave you with a few tidbits in the meantime. :)
There's an interesting new paper here: Algorithmic Decision Making in Financial Services
This paper highlights the correlation between corporate ethics and interests with the normative issues arising in respect of algorithmic decision-making.