AI companies love to tout that their models are approaching—or have reached—PhD-level intelligence. This is blatant nonsensical marketing geared towards an audience that deeply misunderstands what a PhD is and what it takes to get one.
This was so interesting to read. I am nine years post my PhD currently.
I agree with your reasoning. AI models can process information and generate text well, but they fall short of PhD-level research because they lack true reasoning, self-reflection, and the ability to build and update a deep understanding over time. They don’t form their own research questions, adapt to new knowledge, or work effectively in areas with little existing data.
While there were moments that it was enjoyable, like breakthroughs in research or deep intellectual discussions, most often I would describe getting a PhD as an intense, stressful, and heavy process. It requires sustained focus, long hours, and navigating uncertainty, which can be mentally and emotionally demanding.
Why do it, then? For me, it’s the opportunity to pursue questions I’m deeply passionate about, to contribute new knowledge to the field, and to grow both intellectually and personally. Even though it’s challenging, the sense of purpose and the small victories along the way make it worthwhile.
Seriously why not just work half the time or work when you want? Why dies it necessarily have to be stressfuo or long hours? I don’t see why it shouod be or has to be? And surely you can set your own scneduoe. I suffer from depression and don’t need any stress or deserve any more unenjoyable.things on my life.
Thanks for answering but why does it mecessarily hqve to be stressful.and long hours. Couldnt youbhave made it less stressful and little to no long hours?
So spot on it made me laugh. All of the weird peculiarities of the systems you mentioned I’ve experienced. The one about changing one part of a figure only to break something else use to drive me crazy. I tried to replace graph pad prisms with these systems.
Did you really extrapolate just from reading second hand accounts of the failure nodes? If so, very impressive, as the accuracy of your anecdotes of Claudius are uncanny
It's a mix of reading other people's experiences and trying things on my own. I don't use AI a lot because it usually doesn't take me long to get super frustrated, but I have enough experience with it that I can describe what I expect to happen.
This was so fun to read. Even asking an LLM to give you a lit review produces an intern level of research. It doesn't actually make research easier, like a good graduate RA would do
This article is spot-on in so many ways. Yes, great PhDs are feats of exceptional intellect and resilience. And no, AI does not offer “PhD-level intelligence.”
"if an AI can successfully attend a PhD program I’m willing to concede that this AI likely displays AGI."
Would this say more about the arrival of AGI, or the current design of our PhD programs?
Also, I would like to thank you for your cowplot package and Fundamentals of Data Visualization, which have been enormously influential for every plot I make and analyze since graduate school.
Also helps to have a committed supervisor-can make the difference between success and failure (or more likely to enable the student to handle the inevitable fallow periods without going off the rails).
It is impossible for someone to lie unless he thinks he knows the truth. Producing bullshit requires no such conviction. A person who lies is thereby responding to the truth, and he is to that extent respectful of it. When an honest man speaks, he says only what he believes to be true; and for the liar, it is correspondingly indispensable that he considers his statements to be false. For the bullshitter, however, all these bets are off: he is neither on the side of the true nor on the side of the false. His eye is not on the facts at all, as the eyes of the honest man and of the liar are, except insofar as they may be pertinent to his interest in getting away with what he says. He does not care whether the things he says describe reality correctly. He just picks them out, or makes them up, to suit his purpose.
We don’t collapse under heavy cognitive load. We collapse under the weight of unclear systems.
Your framing cuts straight to the real signal, success isn’t raw brilliance. It’s structured thinking, clear anchors, and small repeated proofs.
Academic journeys thrive on consistency over grand intellectual stunts. The EU code of ethics echoes this by emphasizing gradual, tested contributions over unverified leaps. It’s a simple principle, but it holds the mind steady when identity starts to blur.
Your focus on clarity, rhythm, and honest self-assessment is more than academic advice. It’s a survival strategy for any high-stakes mental journey.
Are we ready to choose simplicity over self-imposed chaos? That’s where the real transformation happens.
As someone who decided not to pursue my neuroscience PhD this year (after wracking my brain, sobbing for hours, wondering if that made me a “bad scientist”)…this piece is an eye opener. All I can say is wow.
Is getting a PhD mostly a fun and enjoyable process? If not why do it? How do you endure a lot of failure and disappointment? And does that necessarily have to be the case?
Thank you for writing about this. It reinforces my observations on how public opinion is much more open to criticizing the industry, i.e. finding this post via organic engagement vs. someone I already follow. I finally feel validated after a decade of being the #1 hater of Musk et al., however unfortunate the circumstances and timing.
The barriers to higher education are intersectional. This concept of intelligence only serves as confirmation bias for whomever is creating, implementing and benefiting from its standardization. Testable, not falsifiable. I hope we can eradicate the zero sum backbone that props up the powerful few.
Even if it can’t do the work to “become” a PHD student, it is as functionally useful to me as a PHD student, whether it actually got the degree or not.
If a company can have an army of unlimited PHDs in any field, that’s hugely useful even if they can’t recreate novel scientific experiments.
This was so interesting to read. I am nine years post my PhD currently.
I agree with your reasoning. AI models can process information and generate text well, but they fall short of PhD-level research because they lack true reasoning, self-reflection, and the ability to build and update a deep understanding over time. They don’t form their own research questions, adapt to new knowledge, or work effectively in areas with little existing data.
Is getting a PhD a mostly fun and enjoyabke process, if I may ask, at least for you? If not why do it?
While there were moments that it was enjoyable, like breakthroughs in research or deep intellectual discussions, most often I would describe getting a PhD as an intense, stressful, and heavy process. It requires sustained focus, long hours, and navigating uncertainty, which can be mentally and emotionally demanding.
Why do it, then? For me, it’s the opportunity to pursue questions I’m deeply passionate about, to contribute new knowledge to the field, and to grow both intellectually and personally. Even though it’s challenging, the sense of purpose and the small victories along the way make it worthwhile.
Seriously why not just work half the time or work when you want? Why dies it necessarily have to be stressfuo or long hours? I don’t see why it shouod be or has to be? And surely you can set your own scneduoe. I suffer from depression and don’t need any stress or deserve any more unenjoyable.things on my life.
Thanks for answering but why does it mecessarily hqve to be stressful.and long hours. Couldnt youbhave made it less stressful and little to no long hours?
So spot on it made me laugh. All of the weird peculiarities of the systems you mentioned I’ve experienced. The one about changing one part of a figure only to break something else use to drive me crazy. I tried to replace graph pad prisms with these systems.
Did you really extrapolate just from reading second hand accounts of the failure nodes? If so, very impressive, as the accuracy of your anecdotes of Claudius are uncanny
It's a mix of reading other people's experiences and trying things on my own. I don't use AI a lot because it usually doesn't take me long to get super frustrated, but I have enough experience with it that I can describe what I expect to happen.
rip claudia this was so funny 😆
This was so fun to read. Even asking an LLM to give you a lit review produces an intern level of research. It doesn't actually make research easier, like a good graduate RA would do
This article is spot-on in so many ways. Yes, great PhDs are feats of exceptional intellect and resilience. And no, AI does not offer “PhD-level intelligence.”
"if an AI can successfully attend a PhD program I’m willing to concede that this AI likely displays AGI."
Would this say more about the arrival of AGI, or the current design of our PhD programs?
Also, I would like to thank you for your cowplot package and Fundamentals of Data Visualization, which have been enormously influential for every plot I make and analyze since graduate school.
Also helps to have a committed supervisor-can make the difference between success and failure (or more likely to enable the student to handle the inevitable fallow periods without going off the rails).
It is impossible for someone to lie unless he thinks he knows the truth. Producing bullshit requires no such conviction. A person who lies is thereby responding to the truth, and he is to that extent respectful of it. When an honest man speaks, he says only what he believes to be true; and for the liar, it is correspondingly indispensable that he considers his statements to be false. For the bullshitter, however, all these bets are off: he is neither on the side of the true nor on the side of the false. His eye is not on the facts at all, as the eyes of the honest man and of the liar are, except insofar as they may be pertinent to his interest in getting away with what he says. He does not care whether the things he says describe reality correctly. He just picks them out, or makes them up, to suit his purpose.
Frankfurt, Harry G., On Bullshit
We don’t collapse under heavy cognitive load. We collapse under the weight of unclear systems.
Your framing cuts straight to the real signal, success isn’t raw brilliance. It’s structured thinking, clear anchors, and small repeated proofs.
Academic journeys thrive on consistency over grand intellectual stunts. The EU code of ethics echoes this by emphasizing gradual, tested contributions over unverified leaps. It’s a simple principle, but it holds the mind steady when identity starts to blur.
Your focus on clarity, rhythm, and honest self-assessment is more than academic advice. It’s a survival strategy for any high-stakes mental journey.
Are we ready to choose simplicity over self-imposed chaos? That’s where the real transformation happens.
As someone who decided not to pursue my neuroscience PhD this year (after wracking my brain, sobbing for hours, wondering if that made me a “bad scientist”)…this piece is an eye opener. All I can say is wow.
Is getting a PhD mostly a fun and enjoyable process? If not why do it? How do you endure a lot of failure and disappointment? And does that necessarily have to be the case?
It's Type 2 fun.
What does that mean? Hopefully not like Type 2 diabetes fun?
Some experts, PhD-level and otherwise, have an annoying tendency to opine as experts in other fields.
Thank you for writing about this. It reinforces my observations on how public opinion is much more open to criticizing the industry, i.e. finding this post via organic engagement vs. someone I already follow. I finally feel validated after a decade of being the #1 hater of Musk et al., however unfortunate the circumstances and timing.
The barriers to higher education are intersectional. This concept of intelligence only serves as confirmation bias for whomever is creating, implementing and benefiting from its standardization. Testable, not falsifiable. I hope we can eradicate the zero sum backbone that props up the powerful few.
Even if it can’t do the work to “become” a PHD student, it is as functionally useful to me as a PHD student, whether it actually got the degree or not.
If a company can have an army of unlimited PHDs in any field, that’s hugely useful even if they can’t recreate novel scientific experiments.
An interesting and absolute on point read
Several of my neighbors have Post Hole Diggers, really only good for thoughtful fencing.