D. Justhy's Blog

"Getting to Yes, Now!"

We Need to Become an Ethical Society Before AI Becomes Mainstream

A few months ago, researchers at Stanford University created quite a stir by testing an Artificial Intelligence (AI) program that could detect and identify homosexuals. The program used a machine learning algorithm to look at images and guess who is more likely to be a homosexual.

Now, despite the researchers claiming that their intentions were good (they wanted the AI to help in protecting homosexuals), the study attracted major criticism. LGBT advocates such as GLAAD and the Human Rights Campaign came out strongly against the idea of what is now jokingly being referred to as “the gaydar.”

They believe that the AI can be misused by governments to actually persecute homosexuals, which, to be honest, isn’t hard to believe.

There’s a bigger problem

Using AI to identify gay people is a problem alright, but, this particular case points to something even more insidious – the sheer lack of ethics when it comes to matters involving AI. Even at the research level, there is no ethical framework to guide researchers on AI-related projects. Rules are often made up ‘on the go’ and naturally, this problematic attitude has leaked onto the commercial side of things as well.

In a world where AI is beginning to seep into our daily lives, the ethical aspects of this major transition are growing to be a major concern. In fact, the problems are already at our doorstep.

When robots, bots and AI become mainstream, replacing roles (not jobs!) are people adequately trained and educated to cope with this dramatic shift? In case of driverless vehicles for example, how are we going to deal with the dilemma of having to choose between having people employed and reducing human error on the road?

The need for ethics

The problems with AI have very little to do with the actual technology, itself; they have more to do with the ethical systems that seem absent. We need measures in place to establish ethical frameworks that can guide researchers, product and service creators. Other than that, we also need to have more consensus on AI-related matters from the wider society as well as our governments.

Only then can AI prove to be as fruitful as advertised.

Ethics has played a primary role in all human endeavours and to neglect its function within the area of Artificial Intelligence would be foolish.