Elite Colleges Will Survive AI. The Rest Need to Adapt.

Elite colleges will survive AI due to prestige and IQ-filtering admissions, but middling institutions face existential threats from AI-enabled cheating, which devalues credentials. The status quo tends to adapt and persist, as seen in the resilience of human writers and blogging despite AI advances.

I have long argued https://greyenlightenment.com/2026/05/21/ai-will-not-hurt-the-signaling-value-of-elite-colleges/ that AI will not hurt the status/prestige of elite colleges. This is because the difficulty and selectivity implicitly act as an IQ filter. Even if AI enables widespread cheating in coursework, admissions can still be based on assessments that are largely immune to AI, such as proctored SATs and mathematics competitions, where the use of electronic devices is prohibited or limited. So, even with AI, getting into Harvard still signifies above-average IQ. Detecting AI is especially important for middling colleges, that cannot rely on prestige alone to differentiate themselves. If cheating becomes too widespread, the credential will become devalued in the eyes of employers, leading to reduced attendance and wasted money. At worst, it could pose an existential threat to the institution of higher education itself, save for elite colleges. Employers would be forced to rely less on degrees to prescreen applicants. But schools are fighting back against AI, leading to an arms race between students and teachers. The same https://www.reddit.com/r/cscareerquestions/comments/1qxttnb/any stories of catching someone using ai/ is seen between job seekers who use “interview assistants” to help with coding or other technical questions, and employers relying on increasingly advanced software to detect it. This has led to online complaints from applicants who claim their job offers were rescinded after their cheating software was detected, despite manufacturers’ promises that it was “undetectable.” The thing is, I have found that the most predictable or obvious outcome tends to be the least probable one. This is why I am more optimistic about the viability of higher ed despite AI. Think back to all the failed predictions since 2023 of an “AI bubble” or “mass unemployment” due to AI the US unemployment rate remains historically low, around 4.5% . Or predictions of hyperinflation or deflation due to AI CPI instead has hovered between 2-3%/year . Predictions of obsolesce underestimate how hard the status quo fights back or adapts. The default outcome is that the status quo tends to win, due to inertia and the fact there is a lot of money and careers on the line to keep it that way. A favorite example I like to give is the failure of AI to obsolete writers. One would naively assume that LLMs, by their very affinity for language, would obsolete writers before any other profession. LLMs first demonstrated their capabilities on writing tasks, before moving to coding and math. Yet the opposite has happened, with coding being overtaken by AI, but not writing. AI-generated writing remains clunky despite multiple generations of increasingly capable models, and it still exhibits telltale stylistic quirks it just sounds “off” . As a result, skilled human writers remain in demand. Blogging as a medium is thriving despite AI and LLMs. It has moved to Substack + Twitter and newsletters, and away from Google and Facebook as a source of traffic generation. Many people are easily making six-figures on Substack now, and also combined with Twitter monetization. Neither of these existed 5 years ago. There are way more blogs now compared to 2005-2015 or so, and much longer and technically proficient writing compared to the terse blog posts that dominated 1-2 decades ago. I’m old enough to remember when the only “long form” blogging in town was Scott Alexander’s blog, and the occasional Gladwell article, who originally wrote for The New Yorker . The New York Times and The Atlantic have pivoted to covering everything pertaining to AI. Instead of being behind in the conversation, they are leading it. But to bring this back, the onus is on higher ed to adapt to AI, not to force students to not use AI, similar to how colleges adapted to Google in the early 2000s. They couldn’t prohibit students from using the World Wide Web for help with assignments. Likewise, employers will need to adapt too and look beyond degrees.