I am contemplating whether or not to do a PhD in ML(received an offer from a good North American lab). Financially it makes no sense. However, I am wondering whether it opens up new possibilities in career advancement in the industry. The reasoning is that, scientists get more autonomy and hence a greater chance to make an impact. Is this the case in the real world?<p>edit: I understand that some can work on a PhD without any financial motivation but purely as a passion. I enjoy doing research just for the sake of it but I do not have the freedom to not think about finances.
I have a PhD in SE and it didn't help me financially. But it was an awesome experience in research, learning and personal development. My take is that there's far more to life than chasing after money. Living within your means is an easier path to happiness.<p>To answer your question: If you can potentially land a job at a FAANG or Microsoft Research or in SV or NY, then it probably will pay-off in the long run. Academic research jobs are very hard to get. By the time you graduate with a PhD, ML might no longer be the trendy field. AI has been on and off trend for over 50 years.<p>In summary, whilst there is a chance of eventual financial advantage, the real value of a PhD is in being paid to pursue an interest that you are passionate about.
I think your question is based on a false premise that scientists get more autonomy and more autonomy means a greater chance to make impact. Just getting a PhD would not gaurentee either or these. To make deep impact in industry you would have to assemble or work with "A" grade teams. Academic PhDs seldom do that, more often than not PhD is an individual effort. Some skills might transfer but PhD is typically not structured to impart skillset that would directly contribute to growth in industry. I would personally think getting into industry after Master's is a better choice.