An anxiety-ridden robotic companion for a group of space explorers. A nefarious computer whose malfunction results in the murder of the entire crew on a spaceship. An intimate partner whose entire existence was fabricated in a lab. These are just a few examples of the way artificial intelligence, or AI, has been represented in popular culture. While perhaps science fiction writers’ imaginations are rather distant from our present reality, the rise of AI in our modern society has resulted in massive changes to the way we conduct our lives. With the recent release of user-friendly generative AIs to the public, such as ChatGPT and DALL-E, it has become clear that AI is shaping our future, and therefore we need to be having discourse around what future we want AI to create for us. This issue was the highlight of the 2023 installment of the Kingman Lecture Series held at Andrews University. This year’s featured speaker was Dr. Dewey Murdick, executive director of the Center for Security and Emerging Technology at Georgetown University. At his lecture, held on Wednesday, October 18, Dr. Murdick gave an overview of where AI is today, where it’s going, and what concerns we should have when technology progresses. In addition to being very informative, Dr. Murdick’s presentation shed much light on the ways we should handle AI so as to promote the flourishing of humanity.
Dr. Murdick’s lecture began with a general description of what AI is. He explained that the term “artificial intelligence” is actually very vague, and can be applied to nearly any form of modern technology. However, the AI that is getting a lot of attention in recent times is largely that which falls under the category of deep learning. Deep learning is a type of machine learning, which occurs when computers are used to recognize statistical patterns in a data set. Deep learning uses a mechanism called a neural network, which is a set of interconnected nodes, with varying weights given to each connection. The structure mimics some neuroscientists’ theories on how the human brain works, and is able to analyze inputs for patterns in a way that often appears similar to human intelligence. But all of this begs the question: what is intelligence? Dr. Murdick elaborated by saying that much of the artificial intelligence in existence today exhibits narrow intelligence. This is a format in which a computer can do one and exactly one task that it was designed to do. Broad intelligence, or the ability to do many varied, seemingly unrelated tasks, is still being researched. In general, machine learning is accomplished by having the computer give random answers at first. Then, the researchers give the computer feedback based on the accuracy of the answers, and adjust the model accordingly. This allows the computer to find patterns that humans aren’t even aware of, and sometimes use those patterns to create new products.
One of the subsets of AI to which Dr. Murdick gave special attention is generative AI. Generative AI is AI that “produces content” — think chatbots like ChatGPT or art bots like DALL-E. These AIs have been making steady progress with the rest of technology. One thing Dr. Murdick mentioned regarding this improvement is the way the greater volume of information and data on the internet, working in tandem with the increase in computing power, is allowing for much more accuracy in AIs’ ability to make predictions, particularly predicting the next word in a sentence. This ability has a wide array of uses, from chatbots that simulate human speech to writing assistants like Grammarly. But AI is making leaps in far more spheres than just the office. Dr. Murdick gave examples of AI’s progress in a wide array of fields, and mentioned that advances in medicine and military technology look particularly promising. But with progress inevitably comes a slew of problems. The next portion of Dr. Murdick’s presentation focused on the risks associated with the modern capabilities of AI.
One of the more memorable analogies of Dr. Murdick’s presentation likened AI to an improv artist. In an improv group, no idea is too ridiculous, too silly, or too unrealistic. As a matter of fact, the sillier and more ridiculous the concept, the more valuable the entertainment. AI displays a similar level of detachment from reality. Since AI is nothing more than a complicated algorithm, it doesn’t have an understanding of human ethics, the multifaceted nature of most human decisions, or many other important aspects of existing in our world. The lack of “common sense” within AI can lead it to make questionable decisions. In addition, all AI relies on an exorbitant amount of training data. When these data display biases, the AI learns to discriminate. This is called data-driven bias, and it is one of researchers’ concerns about AI. Another big issue is the potential for deception. As generative AI becomes more powerful, the ability to create “deepfakes,” or highly realistic fake media, becomes more and more accessible to all people, including those with sinister intent. While a photo of the pope in a white puffer jacket might be amusing, there are far darker possibilities of what this technology could achieve.
After addressing all these concerns we should have about the power of AI, Dr. Murdick attempted to end on a hopeful note by explaining what we can do in order to stay ahead of the curve. While many claim that AI will cause the decline of civilization, this is not a new argument. The same was made for technological advances like television and the Internet. As these advancements from days gone by have shown, technology is fundamentally a tool. It has the power for either good or bad; what matters is how we use it. AI isn’t going away. But we can learn how to work with it, in order to prevent it from destroying us. Dr. Murdick said that one of the most important aspects of learning how to live with AI is human-machine teaming. The lack of grounding in the real world means that allowing AI to work without supervision is probably not a good idea. As AI continues to develop, it is of utmost importance that we work alongside the technology in order to curb potential problems before they get out of hand. Dr. Murdick made clear that we need a framework for reporting incidents related to AI. This will ensure that tech companies are held accountable, and technology is being used safely and ethically. Also, Dr. Murdick suggested that young professionals attempting to remain marketable should lean into the most human of our skills — creativity, curiosity, and interpersonal interaction. At the end of the day, humanity is far from being replaced. We just might need to learn how to be flexible.
All in all, Dr. Murdick gave a fascinating and accessible presentation on the implications of a world with artificial intelligence, and how we can channel this new tool in the best possible way. From the picture Dr. Murdick painted, our world is pretty far from the science fiction dystopias of popular media. But Dr. Murdick challenged the Andrews community to be flexible, and to continue to learn how to be world changers in a changing world.
The Student Movement is the official student newspaper of Andrews University. Opinions expressed in the Student Movement are those of the authors and do not necessarily reflect the opinions of the editors, Andrews University or the Seventh-day Adventist church.