Recently, the news has been filled with viral reports such as “Google’s AI wins Gold in IMO” or “AI beats Top Human Performance in Math Competition.” For context, IMO is the International Math Olympiad, a rigorous and challenging math olympiad in which top contestants from each country compete. It is extremely time-consuming. Many people may be wondering how these recent developments in the news will impact their lives. For one thing, should we students even be slaving away doing hours of practice for something that AI could answer for us in a few minutes?
For those of us who invested a lot of time and energy in competitive math, the rapid development of AI really threw a wrench in our plans. The fact that AI can now solve complex Olympiad problems or coding tasks that even top humans take hours to solve within a few minutes indicates that we should be shifting our priorities. Before the rise of ChatGPT and Grok’s “expert modes,” it was perfectly viable for students in math competitions to pursue what was known as the AMC track. They would take a whole year to prepare for the AMC10/12 and AIME competitions, and hopefully advance to JMO/AMO to receive some significant awards, which would prove their mathematical prowess. However, as digital test-taking has become more common and AI chatbots are becoming more powerful, this is no longer a viable option. Just this year, there was a huge increase in the number of students who have made AIME and JMO, rapidly inflating the cutoffs for the Mathematical Olympiad Program, a prestigious nationwide summer camp. It is very plausible that determined and strong-willed individuals can drastically improve their problem-solving skills in a single year, yet the probability of all students increasing so fast is near zero. This sudden jump is most likely attributable to increased cheating through leaks or the use of AI. It is also worth noting that this year, the overall performance on secure, paper-based math competitions did not correspondingly rise with AMC’s; in fact, the average performance on such competitions has fallen in the past two or three years. This means that the new generation of mathletes likely consists of a large portion of cheaters who rely on AI to solve their problems, which is rather alarming and concerning.
This increase in AMC and digital math competition scores means that AMC and other institutions should step up to increase testing security. For one thing, if the Mathematical Association of America (MAA) actually cracked down on test security, leaked AMC problems circling around on Art of Problem Solving (AoPS) or Reddit wouldn’t even exist. Additionally, proctors should actually monitor the students while they are taking the exam, or the AMC organization should have students install some software like LockDown browser that would prevent students from accessing other apps during the exam. There have already been multiple reports that students taking the AMCs on a computer directly ChatGPT’d the problems; the proctors didn’t notice because they were on their phones. After all, the proctors were all unpaid volunteers who had no interest in actually ensuring the academic honesty of the students. Ironically, schools and educational institutions like MAA are always harping on testing integrity, yet there has been no action from them regarding the recent cheating incidents.
Numerous students were negatively impacted by the large influx of people who were using AI in math competitions. Many who took math competitions very seriously suddenly felt huge burdens on their shoulders due to the artificially inflated cutoff lines, and in order to prove their authenticity, they would flock to other paper-based, onsite math competitions, greatly increasing the competition and pressure on everyone involved. But this leads to another good question: After ChatGPT has already demonstrated the ability to do math competition problems far more efficiently and cost-effectively than any human, why do so many people still participate in math competitions? The answer would be that ever since the buildup of mounting college pressures in the early 21st century, students have been looking for various outlets to showcase their academic excellence, and math competition happens to be one of them. From the artificially nuanced and overcomplex math competition problems that would never appear in the real world, it is already quite clear that math competition is no longer about problem-solving skills; it is simply a rat race in which students attempt to prove their IQ to the college admission officers. Perhaps it’s time for students to reconsider.
