“Don’t Even Bother Getting a Law or Medical Degree"

“Don’t Even Bother Getting a Law or Medical Degree”

This was quoted from an interview by Jad Tarifi, founder of Google’s GenAI team. He went further to say, “Even things like applying AI to robotics will be solved”. I do not doubt this person is clever and at the forefront of their field, but I want to remind you that Lord Kelvin, the physicist who invented standard temperature measurement and thermodynamic laws in the 19th century, once said “There is nothing new to be discovered in physics now”. What leads to such hubris is a mystery to me, but my guess is that these people operate in bubbles with compounding and frequent successes blinding them to other failures or shortcomings.

For those of us practising in the field, the problems are clear. For example, I personally know a doctor who has embraced AI. He uses it for transcribing and summarising his dictation and consultation (with his patients’ consent). However, complex and emerging medical terms are often not identified, patients do not always speak clearly and go off on long tangents confusing the AI. Similarly in law, when AI is used for case law it will hallucinate legal precedents, and if these are presented in court, a lot of credibility is lost by the lawyer.

The problems in robotics are even worse, not only do the latest fleet have prohibitive costs for mass adoption but trying to navigate dynamic unstructured environments such as people’s homes, in a safe way, is an AI’s worst nightmare. The videos that are emerging of the latest robots have one of two things happening: 1. It’s human remote controlled or 2. It’s cherry picked from a myriad of failed runs.

Let’s say, AI no longer hallucinates or forgets information and can take humanoid form at a low cost. Something rarely considered is empathy and wisdom. This counts for a lot in medical prognoses, client relationship and product design. Much like current robotic customer service agents, people are going to be asking “can I see a human doctor” 9 times out of 10.

I would go so far as to argue AI will make human expertise more valuable. I’m seeing this in real-time in software engineering. Programmers who thoroughly understand code and software systems are pulling ahead of those that don’t because there’s so much broken AI code to make sense of that a human touch is a necessary component for success.

So, whenever I’m asked by a high school graduate if they should even bother studying computer science or data science. The answer is absolutely yes. I think the same should be said for aspiring doctors and lawyers.

1 September 2025