pull down to refresh

While other new students fretted over the university’s rigorous core curriculum, described by the school as “intellectually expansive” and “personally transformative,” Lee used AI to breeze through with minimal effort. When I asked him why he had gone through so much trouble to get to an Ivy League university only to off-load all of the learning to a robot, he said, “It’s the best place to meet your co-founder and your wife.”
To be clear, Lee doesn’t think this is a bad thing. “I think we are years — or months, probably — away from a world where nobody thinks using AI for homework is considered cheating,” he said.
In late March, Columbia suspended Lee after he posted details about his disciplinary hearing on X. He has no plans to go back to school and has no desire to work for a big-tech company, either. Lee explained to me that by showing the world AI could be used to cheat during a remote job interview, he had pushed the tech industry to evolve the same way AI was forcing higher education to evolve.
Lee has already moved on from hacking interviews. In April, he and Shanmugam launched Cluely, which scans a user’s computer screen and listens to its audio in order to provide AI feedback and answers to questions in real time without prompting. “We built Cluely so you never have to think alone again,” the company’s manifesto reads. This time, Lee attempted a viral launch with a $140,000 scripted advertisement in which a young software engineer, played by Lee, uses Cluely installed on his glasses to lie his way through a first date with an older woman. When the date starts going south, Cluely suggests Lee “reference her art” and provides a script for him to follow. “I saw your profile and the painting with the tulips. You are the most gorgeous girl ever,” Lee reads off his glasses, which rescues his chances with her.
Before launching Cluely, Lee and Shanmugam raised $5.3 million from investors, which allowed them to hire two coders, friends Lee met in community college (no job interviews or LeetCode riddles were necessary), and move to San Francisco. When we spoke a few days after Cluely’s launch, Lee was at his Realtor’s office and about to get the keys to his new workspace. He was running Cluely on his computer as we spoke. While Cluely can’t yet deliver real-time answers through people’s glasses, the idea is that someday soon it’ll run on a wearable device, seeing, hearing, and reacting to everything in your environment. “Then, eventually, it’s just in your brain,” Lee said matter-of-factly. For now, Lee hopes people will use Cluely to continue AI’s siege on education. “We’re going to target the digital LSATs; digital GREs; all campus assignments, quizzes, and tests,” he said. “It will enable you to cheat on pretty much everything.”
this territory is moderated
A friend of ours used AI to complete most of her coursework and has become a professor who uses AI to generate most of her coursework.
AI making work for AI to do.
reply
Must be a recent phd grad then? GenAI has onlyreally been around for 4 years or so.
reply
Recent, but not in a field or program that requires a PhD for faculty positions.
reply
Hmm, seems a bit disturbing
reply
I wouldn't be surprised to see this sweep through the unnecessarily credentialist industries.
Repealing dumb occupational licensing requirements would probably clean this up very quickly.
reply
I'm guessing this is in some medical or social services adjacent field then?
I was explaining to my wife that industries supported by public money will tend to be full of unnecessarily burdensome occupational licensing. It's part of the problem of how to have accountability when you're spending other peoples' money. IMO the best solution is just to have fewer industries supported by public money.
I think AI usage is going to be assumed, like Google searches can be assumed, and calculators can be assumed, for any take-home assignments.
As professors, we just have to level up our assignments. Make them more project-oriented rather than tests of rote knowledge.
In my class's final project, I let them use AI as much as they want. And basically, the quality of the final project is pretty much consistent with the performance of the students throughout the rest of the class. The bad students struggle to use AI effectively, the good students are using it well and making some pretty nice projects with it.
reply
That sounds like the ideal. In principal, we should now be able to expect better outputs from students.
reply
Yes, making AI usage cheap and easy while making the rewards for using it very high is the perfect incentive for plagiarism!! It makes me stop and think what will happen when businesses and industry catch on to what is going on in the colleges. What they will do? Can you say intellegence, aptitude and knowledge tests? If I were them I would be instituting merit tests to beat the band. Anyone who doesn’t do it will get just what they deserve—probably only well connected idiots!
reply
The best part about this besides for the fact it's gonna be unraveling the college myth, is that by cheating and seeing how the LLM produces the work they're probably learning more than struggling on some esoteric topic and receiving mediocre feedback from a teaching assistant
reply