Artificial intelligence has been a hot button topic for a while now. We’ve written a few articles on it and where we think it’s going. We can tell you that our inboxes are flooded with blogs and news on the topic every day to the point it’s getting a little annoying. That being said, we don’t think AI is going anywhere. In fact, it’s probably just getting started. While we hope it doesn’t take jobs and is used as an added tool in the arsenal for employees, we can’t say for sure how it will pan out. One area of interest when discussing AI is the ethics. As humans, we have cognitive bias from our life experiences, whether we realize them or not. We’d like to think that because AI is machine learning it does not have those biases and that’s what makes it more efficient. However, we must remember that humans program the machine and therefore, must follow certain rules and regulations in order to not infringe upon human rights & civil liberties. The academic community has come up with The Belmont Report as a guide for algorithmic development & experiment.
1. Respect for persons: this focuses on consent and autonomy of individuals to protect them.
2. Beneficence: similar to healthcare ethics “do no harm” in order to keep AI algorithms from amplifying biases from race, gender, etc.
3. Justice: fairness and equality of who reaps the benefits of machine learning.
a. Equal share
b. Individual need
c. Individual effort
d. Societal contribution
e. Merit
There are a few hot issues with AI ethics for now. They include:
1. Impact on jobs: we touched on this above but it happens every time some new tech comes out that will vastly change our lives. We think that roles will shift but not be taken over. There will need to be people who oversee the AI and all the duties it performs. Therefore, one job may turn into another but not disappear altogether.
2. Bias & discrimination: this has happened in the past with AI when training data leads to biases. It’s important, as we stated above, to program the system to be unbiased and not discriminatory. Although difficult, crucial and may take years to refine.
3. Privacy: this is an issue now as it is with the digital world we live in and 3rd party cookies. States are coming up with new laws to protect user data but how will AI deal with personal data? Most users don’t even know that their data is being used to make decisions that may affect them. The answer has yet to be seen.
4. Accountability: there aren’t any legislations yet to control AI practices since it’s so new. This is why ethical guidelines have been created but there needs to be some sort of repercussions for unethical uses. We see this coming into play in the coming years as AI advances.
5. Singularity: think Class of 09’ on Hulu. The scary thought that AI will surpass humans and develop its own thoughts and ideas.
Ethical issues can create problems for businesses, such as legal issues, brand damage, product failures and, in turn, loss of revenue. Understanding AI will be important for everyone in the near future if not already. AI ethics are being taught in some high schools and middle schools for the next generations in order to keep it in check. No brand wants to be the one that accidentally puts out machine learning that is discriminatory or get sued for copyright infringements. In the next few years it will be a balancing act to create the best practices and laws as artificial intelligence grows at a rapid rate.
