Ethics in artificial intelligence (AI) refers to the values that guide the development of smart technologies such as machine translation software, facial recognition applications, and streaming services that use recommendation algorithms. But why should we care about ethical issues surrounding artificial intelligence?
“Because we're all already part of a variety of AI systems, and they're evolving rapidly. We're now wondering if we can identify political opinions from a person's face or not. We discuss whether we can study underlying biases through algorithms. The more widespread and diverse their uses become, the more they will impact us all. We need to feed ourselves, and part of that is understanding the ethical issues.” Anna Mari Rusanen. She is a university lecturer in cognitive science and the coordinating teacher of the new course “Ethics of AI” at the University of Helsinki.
1. Identify the technology to support
Artificial intelligence is already being used everywhere large amounts of data are processed and analyzed. Healthcare experts are debating whether algorithms and data should be used to predict and prevent health risks. Algorithms are already used to calculate taxes and social security. In the future, national security, transportation, urban planning, energy and food production will increasingly involve robotization and the use of algorithms. On social media, algorithms decide what kind of content to show you, and on dating apps, they calculate what types of potential partners should be shown to you.
“Each of these issues involves several ethical factors. We evaluate the benefits and risks of the system, as well as the implications for human rights, etc., and we evaluate the benefits and risks of the system, as well as the impact on human rights etc. “We need to decide who will take responsibility in this case,” Rusanen said.
And when we visit a shopping center, for example, we are constantly monitored, with automatic image recognition systems calculating the visitor's path.
“Many people are agreeing to this for coronavirus tracking purposes, but at what point does automatic aggregation turn into surveillance?”
Rusanen says you need to identify which aspects of the technology require a response.
“Technology itself is neutral, and what's great for one application may not be so great for another. Being able to unlock your phone with your thumbprint is nice, but if fingerprint authentication doesn't work… For example, it is important to recognize that technology violates privacy in ways that are no longer acceptable. Ethical issues are usually not about the functionality of the particular application itself, but about the application and social context of the technology. It occurs from.”
2. AI ethics are changing
Now is the moment to decide what intelligent systems we will use in the future. New technological solutions may even shape the relationship between states and citizens, placing new limits on them, such as through automated decision-making. New civil rights are emerging, such as the right to be forgotten.
“Politicians and those creating actual AI solutions will need to consider surprisingly deep questions, such as the nature of the default public organizing AI-based public services,” Rusanen says.
Politicians are aware of this, Rusanen said, and the EU plans to publish a draft new AI regulation later this year.
“They specify what is and is not allowed in AI systems. The new rules emphasize human rights and suggest what kinds of applications are at high risk of abuse. For example, facial recognition could be an example of such an application,” explains Rusanen.
“It’s time for all employees in government, data processing companies, and the technology industry to wake up and learn the basics of AI ethics.”
3. AI ethics is not a fraudulent accounting process
Rusanen points out that AI ethics is not a window dressing for finished products, services, or systems. For example, when a company uses AI, it is necessary to establish ethical principles at the stage of creating the system.
“Developers need to think about who their products and services relate to, how the rights of these people will be protected, and what consequences they may face. Everyone who works with artificial intelligence needs to be aware of what ethical principles actually mean for their day-to-day work. This needs to happen at each design meeting and at the beginning of the process. “It's not enough for an ethics advisor to approve the final results,” says Rusanen.
According to Rusanen, some Finnish companies want to improve in this area.
“Many companies promote moral regulation within the EU, but have no qualms about selling products to China where they are used for mass surveillance. That's hypocrisy.”
Join us for our AI Ethics course!
read more:
The Ethics of AI online course challenges you to think about what technology should be used for.
