"Should I buy a house now?" "Would it be better to change companies next year?"
Beyond merely searching for information, we are now in an era where people consult AI for major life choices.
Can decisions made by following AI's answers truly be called "our own choices that we made ourselves?"
Professor Philip Koralus, who researches philosophy and artificial intelligence at the University of Oxford in the UK (Inaugural McCord Professor of Philosophy and AI, and Founding Director of the Human-Centered AI Lab at Oxford), describes this future of outsourcing life's choices to AI as "Autocomplete for life."
It is the concept that, just as a smartphone screen predicts and shows the next word we might type, AI presents our next options in life.
Are we in the future actually leaving all sorts of choices to AI?
Koralus worries that if AI replaces human judgment, it could lead to a "loss of autonomy" for humans.
[Philip Koralus / Inaugural McCord Professor of Philosophy and AI, University of Oxford (May 2026, at a seminar co-hosted by the Oxford HAI Lab): To solve the problem of autonomy in the relationship between humans and AI, we must not let AI "autocomplete" our lives.]
What Koralus focuses on is how to help humans think for themselves rather than having AI substitute human judgment.
[Philip Koralus / Inaugural McCord Professor of Philosophy and AI, University of Oxford (May 2026, at a seminar co-hosted by the Oxford HAI Lab): How can we make AI put on the brakes appropriately when needed, while still making us want to keep talking to it?]
At the center of this are "catalytic questions."
For example, instead of letting AI give an answer when asked, "Should I change companies?", the AI should ask questions back, such as "What is the biggest reason you are considering changing jobs?" or "Which choice do you think you would regret more five years from now?"
It is about AI not merely asking many questions, but throwing important and piercing questions that prompt humans to re-deliberate, ultimately helping them reach a judgment they can accept themselves (Remain the authors of our own judgments).
[Philip Koralus / Inaugural McCord Professor of Philosophy and AI, University of Oxford (May 2026, at a seminar co-hosted by the Oxford HAI Lab): It is a very natural principle that "humans must ultimately be the ones responsible for their judgments."]
No matter how advanced AI becomes, what must we do to remain humans who decide and choose for themselves?
Professor Philip Koralus, visiting South Korea, will present the solution at the SBS D Forum on the 5th.
Reported by Jeong Yunsik | Video by Kim Jong-tae | Graphics by Yook Do-hyun | Produced by SBS Digital News
※
Copying, redistribution, and unauthorized use in AI training are strictly prohibited.