Should I pursue a career in Artifical Intelligence?
I am 3 years into a finance graduate programme and bored rigid. I am interested in AI from the point of view of safety and protection from "the singulairty" etc. and have limited coding experience currently although I am highly quantitative as well as qualitative - excuse the horrible terms. #artificalintelligence #computing #finance #careerchange
Hello Ben: That is a great question. From my experience advising students in undergraduate studies, I've seen students change majors often however, I tend to recommend finalize any changes near the end of the second year of studies. You may already be familiar with the reason but in the third year, and sometimes the second, you begin to take course directly related to your major area of study,in a focused area and field of work. As such, often times, these fields are not related to other majors and therefore credit hours, time towards graduation, and money are lost. I would encourage you to assess how many credit hours you have already completed in the direct major of study. As I ready this further, you are in a graduate program which means all of your credit hours are in that major and field. As such, it would be a large sacrifice to foregoe three years of studies and credit hours. Consider these questions: why did you originally choose finance as a major course of study (for yourself or strong influence of others?); What do you dislike about finance? Is it bareable for a few years of work or not at all? Do you have the means and funds to stop now and transition to AI? Do you have the fortutitude to start again? Are you related to anyone working in AI? Trusted associates in AI? Also, review the occupational outlook handbook website to get a greater sense of the innerworkings of employment in the field of AI (try searching computer engineering, or software development) including the demand/need for employees in the field, pay ranges, and average day. You can also try onetonline.org for more research into the field of AI. You may know better search terms to use of you are quite familiar with the field already. Any event, it is a serious undertaking to switch and do contemplate each question provided. You may even write a comparison chart between the two. For a lifetime of satisfaction, change may be worth it but be sure it's for the right reasons (not being influenced by outsiders, strictly money, or for prestige) but rather fulfillment, satisfaction, and contributing to society. Best of luck in your decision!
Lashay recommends the following next steps:
Hey Ben, great question. It's awesome to see more people getting interested and excited about AI, especially around AI safety and ethics. I am an AI researcher (focusing on machine learning and deep learning for natural language problems) at a VC funded startup. There are many different ways to enter into the AI space. My undergraduate was in Creative Writing and the social sciences.
Are you in a doctoral program or a masters program? If you're in a master program, finish it, as any advanced degree is better than no degree. If you're in a doctoral program, see if there is a way a to switch your thesis topic. You also have the opportunity to exit with a masters in passing (assuming you've passed your quals) and it might be worth seeing if that's you want to do.
AI safety and security can be approached from different perspectives. Usually you won't need to write any code for this space. There is the policy side which tends to be less technical. You don't need to know how to code but you quantitative and qualitative background is valuable to be able parse technical literature and think deeply about the unintended consequences of AI technology. The policy side is rapidly evolving and emerging. You can contribute immediately by blogging as a thought leader or looking for opportunities to join legal teams that do risk mitigation. The academic route is going into public policy programs.
The other way is actually doing research in algorithmic safety and detecting latent discrimination in models. There is an entire space of emerging research into investigating the hidden biases and ensuring algorithmic integrity. Most of this work is mathematical and theoretical (usually around statistical sampling, stratification, and analyzing results to detect latent biases).. Most of the work in this space is academic and usually around producing new technique and methods. You can jump in and contribute to the literature by publishing, going to conferences, and joining the research community. If you want to formally enter this space, you'll most need to either join a research lab or look for companies hiring theoreticians for internal model audits. Your quantitative background is a great preparation for switching into a computer science or statistics program where you can focus on this work in a formal academic setting.