The evidence of AI outperforming trained graduates on knowledge-intensive tasks is now well-documented across multiple professional domains.
I did a little research and the results show AI not just outperforming humans in Knowledge but also helping humans outperform others.
GPT-4 has reportedly scored an average of 75% across MCQ-based exams in a variety of disciplines. More specifically, it scored 86% on the US Medical Licensing Exam Step 1, this is a benchmark said to be designed to test high-level problem-solving and the application of core clinical knowledge.
It's also interesting to find that GPT-4 has also passed the entire Uniform Bar Examination, including not just multiple-choice questions but also open-ended essay components and performance tests, the same test that determines who may practice law.
Then out of Spain, report shows that in a rigorous Medical Intern Resident (MIR) examination in 2025, a fine-tuned model achieved 195 out of 210 questions correctly, a score that would place it among the top human candidates.
When we take a step back or away from exams and look beyond, we find a peer-reviewed randomized controlled trial published in Scientific Reports in June 2025 showing that an AI tutor outperformed traditional in-class learning with an effect size between 0.73 and 1.3 standard deviations.
Students using the AI tutor achieved substantially higher post-test scores in less time, a median of 49 minutes versus 60 minutes for in-class learners
The Labour Market Reality Check
All roads from education leads to working till death do the employees and their employers apart right?
It's only expected that this is where the discussion leads.
Jobs, what is the impact?
If AI can pass the bar exam and the USMLE, what is happening to the graduates who sit those same exams?
Well so far, we have data from the Cengage Group's 2025 Graduate Employability Report which reveals that this year's graduates face the most constrained entry-level labour market in five years.
About 76% of employers are hiring the same or fewer entry-level workers, driven partly by the rise of AI (cited by 46% of employers by the way).
So far 33% of 2025 graduates are unemployed and actively seeking work, imagine what that number will look like 5 years from now.
This is happening, right now.
Now I'm sure that if I keep looking, I'll keep finding data that supports the argument of AI surpassing humans in the knowledge field and that milestone is already taking jobs as the tech becomes an integrated infrastructure across multiple domains.
So the question remains, what becomes of education?
There's a consensus of what should come next, folks from research, policymakers, and supposed forward-thinking institutions all conclude on a the direction where we shift our focus from knowledge storage to human value creation.
But what does that even mean?
I'll save you the bullshit.
Education has always been about attaining knowledge to generate value long-term, it had just been mostly branded around working jobs that kills you slowly because that system was beneficial but now it isn't really important so a shift has to take place.
One thing we must understand is that things will always keep shifting. Every job will be about remove "cost layers" and if that happens to include humans, then that's what will happen.
All else doesn't really matter. And in the face of this reality, we will eventually come back to the one thing that determines if a people survive or not and that is their governance systems.
This is what will hold the future together, without the right system, in the long-term, it all goes to shit.
Maybe now the world will see why crypto and blockchain is so important. The future will need decentralized governance. It should sure hope it eats deep into global systems before AI gets out of hand.
Image credit: Digiexam