The ethics of innovation and the future of humanity in generative AI

AI has the power to transform the social, economic, and cultural fabric of the entire world. As television, cell phone, and the internet sparked a mass transformation, generative AI development like ChatGPT creates new opportunities that humanity is yet to imagine.

With great power comes great responsibility. It's not a secret that generative AI raises new questions about privacy and ethics. One of the biggest risks is that society uses this technology irresponsibly. It is vital that innovation and accountability do not overtake each other to avoid this outcome. New regulatory guidance needs to be developed at the rate that tech's biggest players are launching new AI applications.

To fully understand the moral dilemmas surrounding generative AI - and their potential impact on future global populations - we must take a backward step to understand these large linguistic models, how they could create positive changes, and where they might fall short.

Humans answer questions using their genetic makeup (nature), self-learning, education, and observation (nurture). ChatGPT is a machine that has access to all the data in the world. AI's output, just as our biases affect our responses, is biased by the data that was used to train it. Data is often extensive and contains multiple perspectives. The answer that generative AI provides depends on the question you ask.

AI has access to trillions of terabytes, allowing users "focus" their attention by prompting engineering or programming. This allows the output to be more precise. It is not a problem if the technology is being used to suggest actions. However, generative AI is capable of making decisions that can affect human lives.

When using a navigational system, for example, a person specifies the destination and the machine calculates a route based on factors like road traffic data. Would the navigation system's action match that of a human if it was asked to choose the destination? What if a person was not able to intervene, and decided to drive a route different from the navigation system's suggestion? Generative AI simulates human language thoughts based on patterns it has seen before. It does not make decisions or create new knowledge. Legal and ethical concerns arise when the technology is used for this type of use.

If ChatGPT was used in a university class on literature, a professor would be able to use the technology's knowledge and help students discuss the topics at hand, as well as pressure-testing their understanding of the material. AI can be used to support creative thinking, expand students' perspectives and provide a supplemental educational tool. However, this only works if the students have read and understood the material.

Some applications are medium risk and may warrant additional criticism in regulations. However, when used correctly, the benefits can outweigh any risks. AI can, for example, make recommendations about medical treatments and procedures that are based on the medical history of a patient and patterns it has identified in similar patients. A patient who follows the recommendation without consulting a medical expert can have disastrous results. The patient is ultimately responsible for making the final decision, and how they use their medical data. However, generative AI shouldn't be used without proper checks and balanced.

High-risk applications can be identified by the lack of human accountability and AI-driven autonomous decisions. According to our laws, it is impossible for an "AI judge" to preside over a courtroom. Judges and attorneys can use AI for research and to suggest a course of action for the defense or prosecution. But when the technology takes on the role of a judge, it poses a new threat. Judges are stewards of the law, bound to law and conscience -- something AI lacks. AI may have the ability to treat people with fairness and without bias in the future, but for now, only humans are accountable for their actions.

We are now in a critical phase of the regulatory process, where we must consider these applications in practice. While we continue to research AI behavior and develop guidelines, there are four steps that we can take to minimize the immediate risk.

ChatGTP is only the tip of a generative AI iceberg. As technology advances at a rapid pace, assuming responsibility will determine the impact of AI innovations on the global economy and many other outcomes. We are at a unique point in human history, where technology is trying to replicate our "humanness".

technology news