Artificial intelligence researchers are hot in demand. In an increasingly smarter digitalised world, we need more engineers in machine learning, natural language processing, computer vision, and more. But is a PhD really necessary to work in these specialised fields? Or should you jump into the AI market straight away?
There are countless candidates applying to startups built around artificial intelligence. Yet many of the core roles on offer – data analyst, data scientist, robotics scientist – seemingly require a PhD at minimum. This is because there are very few equivalent experiences that can match such rigorous research-intensive projects.
However, it is worth detailing your motivations. If it boils down to a desire for higher potential salaries, that would be a terrible idea to commit yourself to a PhD programme.
“Most PhDs in CS/AI do not end up in a ML Researcher job at a place like Google Brain or Facebook AI lab,” says an industry veteran on social discussion site Reddit. “So there are many people with PhDs from top programs just working as regular data scientists at these companies.”
“It’s arguable whether there will be even a bump at all. A PhD is 5 years of $20K-40K stipend. That’s a lot of lost potential income for half a decade,” chimes in another software engineer.
PhD research in UK
A typical PhD programme in artificial intelligence takes around 3 to 5 years to complete.
In the UK, students tend to aim for the following top universities:
- University of Oxford
- University of Cambridge
- Imperial College London
- University College London
- University of Edinburgh
- King’s College London
- University of Birmingham
- University of Warwick
Deeper academic research around artificial intelligence often face scrutiny from the public. Oftentimes, this is due in part to the unfettered peer review system, since comparable results to which scientists can verify cutting-edge outputs are little to none.
More importantly, we currently can’t decipher just how the artificial intelligence systems reach their conclusions. So this is a serious issue that casts a shadow over the wider computer science community until the “black box” nature of neural networks can be adequately explained.
Pursue Masters, join industry
It all comes down to your values in life. Some students aspire to launch startups that utilise artificial intelligence frameworks that may not involve deeper knowledge besides learning the required tools. PyTorch and Tensorflow are popular programming libraries that require minimal theoretical understanding for instance.
“I believe that building products have a greater contribution to people while conducting research contributes to the knowledge of the corresponding field,” says Raymond Cheng, a Masters student at Carnegie Mellon University. “Although (…) research also contributes to people and society, I think products and apps are a more direct way of providing assistance and convenience to daily consumers.”
[ Stay tuned for our next article as we pose the titular question to graduates pursuing a PhD in artificial intelligence. ]