Stephen Wolfram thinks we need philosophers working on big questions around AI

26.08.2024

Stephen Wolfram, a mathematician and scientist, grew up in a home where his mother was a philosophy professor at Oxford University. This early exposure to philosophy initially led him to steer clear of the subject. However, as he grew older and perhaps wiser, Wolfram came to appreciate the value of deep, philosophical thinking. Now, he seeks to apply that level of philosophical rigor to AI research, aiming to help us better understand the complex issues that arise as AI technology advances.

Wolfram, who was somewhat of a child prodigy, published his first scientific paper at the age of 15 and earned a doctorate from Caltech by the time he was 20. His extensive work spans multiple disciplines, including science, mathematics, and computing. He is the creator of Mathematica, Wolfram Alpha, and the Wolfram Language, a robust computational programming language.

“In addition to fundamental science, my main life’s work has been developing the Wolfram Language, a computational language designed to express ideas in a way that’s useful to both humans and computers,” Wolfram shared with TechCrunch.

As developers and researchers delve deeper into the interplay between computers and humans, Wolfram suggests that this exploration is becoming more of a philosophical endeavor, involving deep reflection on the implications such technology may have on humanity—a form of thinking that is closely aligned with classical philosophy.

“The real question is what you think about, and that’s a different kind of question, one more often explored in traditional philosophy than in typical STEM fields,” he explained.

For instance, when considering how to implement ethical safeguards in AI, these are fundamentally philosophical questions. “In tech, when people suggest that AI should just ‘do the right thing,’ it begs the question: ‘What is the right thing?’” Determining moral choices is inherently a philosophical challenge.

Wolfram has had some “disturbing conversations” with companies deploying AI without sufficient consideration of these issues. “Engaging in a Socratic dialogue about these topics, you’d be surprised at how little clear thinking there is around these issues. I don’t have all the answers—that’s the challenge—but it underscores the relevance of these philosophical questions today,” he noted.

He pointed out that scientists generally struggle with philosophical thinking. “One striking observation is that when scientists are confronted with radically new ideas, it often disorients them because science typically progresses incrementally, rather than through paradigm shifts,” he said.

If philosophy’s primary role is to tackle big existential questions, Wolfram believes we’re entering a golden age of philosophy due to the growing influence of AI, which is raising numerous fundamental questions.

“I find that philosophers are often more flexible when it comes to thinking about different paradigms,” he added.

During a discussion with a group of master’s philosophy students at Ralston College in Savannah, Georgia, Wolfram explored the intersection of liberal arts, philosophy, and technology. He mentioned that he had recently revisited Plato’s “Republic” to reconnect with the roots of Western philosophy.

“The question of how AIs should govern, if they ever do, and what that governance should look like is a modern extension of political philosophy that goes back to foundational questions Plato addressed,” Wolfram remarked to the students.

Rumi Allbert, a student in the Ralston program with a background in data science who also attended Wolfram’s Summer School, was intrigued by Wolfram’s philosophical interests.

“It’s fascinating that someone like Dr. Wolfram, so advanced in his field, sees the importance of philosophy and a humanistic approach. It suggests that his work has evolved into addressing more philosophical questions,” Allbert observed.

Wolfram, having been at the forefront of computer science for decades, is now recognizing the links between philosophy and technology. This may signal that it’s time to approach AI-related questions from a broader perspective than purely technical or mathematical, and incorporating philosophers into the conversation could be a significant step in that direction.