0

I've recently become very interested in the potential AI holds for the future of society. I believe it has the potential to truly alter the way we live our lives in the not-too-distant future. I've read around three dozen books by leading scholars and I've written my undergraduate thesis on the implications of AI for modern warfare and the international balance of power. With that in mind, I've started to think about a career related to AI and machine learning.

However, I have a bit of a problem considering my background as a humanities person. I'm about to finish my undergraduate studies in International Relations and I plan to get a master in a similar field. As someone with little background in Comp Sci/ Math/ Physics/ etc, how could someone go about working in a field like AI? I struggle to see what utility non-STEM people could offer to firms.

5
  • 3
    I suppose you could work in the field of ethics and other stuff in AI, but if you want to be technical I'm afraid you'll need to study a bit longer.
    – Maxime
    Oct 22, 2021 at 13:16
  • 1
    Career advisors at your institution can help with that. You need a long conversation about what your goals are. Just a note; experiencing the benefits of AI and working on AI is about as different as eating a cake and baking it. A lot of people think they like making video games because they like playing them, but nothing can be further from the truth. IMO Video games are the most difficult type of software to create well, and success even with a huge budget is not guaranteed.
    – Nelson
    Oct 22, 2021 at 13:31
  • It's a lot easier for a humanities major to learn about computing than it is for a computer nerd to learn about humanity. Oct 23, 2021 at 22:03
  • "I believe it has the potential to truly alter the way we live our lives in the not-too-distant future." - you should examine more about what AI is and what it is not. Putting too much hope into a system that does "just something" that one can hardly define in advance or trace afterwards can indeed alter our future - but not necessarily to the good.
    – puck
    Oct 28, 2021 at 5:57
  • You may need to consider going back to school to get an MS, if not a PhD, to get the necessary prerequisite work in AI. Keep in mind that AI, and all its associated subfields like ML and NLP are broad and constantly evolving fields, there is still a lot to discover about what exactly these fields are.
    – Daveguy
    Oct 28, 2021 at 18:18

5 Answers 5

9

It's never too late to change your field, as long as you're willing to put in time and effort to learning.

Some of the best people working across an variety of different areas of IT have no formal background in it. Some of them have degrees in unrelated subjects including medicine, music and English literature. Some of them left school at 16 and have no further education.

If you really want to get into the more technical areas of AI, then start learning more about the subject, and also get some programming knowledge or experience. There are tons of free resources available - and if you can demonstrate a decent level of self-learning, and real interest and passion, good companies won't care that you don't have a computer science degree.

2
  • 3
    "It's never too late to change your field, as long as you're willing to put in time and effort to learning." instant like
    – PowerCat
    Oct 22, 2021 at 14:56
  • 2
    AI is not IT. AI is not CompSci. Many companies will hire a big set of interrelated backgrounds (math, physics, econometrics, statistics, compsci, engineering), that have one commonality: A strong STEM background. You might have touched on predicate logic, but without programming experience or many hours spent wrestling mathematical functions in any context, your main disadvantage will be the number of hours required to ingrain a certain way of thinking and ability to do symbol manipulation. Do you have the time to become good enough?
    – MvZ
    Oct 28, 2021 at 20:17
7

Sure, if you're willing to accept not working on the AI directly.

AI companies are like any other tech company- they need people other than developers. They need marketers, salespeople, client managers, mid- and upper-level management staff, etc. It's quite possible that someone with a degree in the humanities might be able to find a job with them working in one of these sorts of business roles.

6

I don't think you know what AI really is. AI, as we have it today, isn't about logic, reason, thinking, or philosophy. Nothing we have is capable of that, and likely won't be in the next 20 years. It may never be- the idea that all you need is sufficiently large processing power to obtain something like consciousness is a supposition that may or may not be true. What we have is pattern matching. You put in a metric ton of data, it analyzes it, and it pops out a small set of rules for categorizing new input based on the original set of input. That's all it does. It isn't trying to understand those patterns- it isn't capable of understanding patterns. It can just find them, and evaluate new data based on them.

Just to drive this home- those rules it pops out, it doesn't understand them. It doesn't know what the different properties are. It basically knows "If you multiply column A by 1.23 and add column B and the result is greater than 5, then the output should be 1. Otherwise its 0" It doesn't know what A and B are. It doesn't know what the output means. It just knows when it does that math, its right the majority of the time.

As such, your humanities background isn't going to actually help. You can write papers on the vision of how AI can be used, and you might get paid for that for publicity at a Google or something. But they'll ignore it in favor of maximizing profit. But you're not going to be involved in building the AI unless you're in programming or project management. You can work at an AI company in some other role of course- those companies need accountants, sales people, HR, etc just like everyone else.

1
  • AI as a broad subject is about logic, reasoning and the like. Understandably, though, ML is somewhat conflated with AI at present because of its current popularity and widespread (mis)use. What you write is largely true of ML.
    – J W
    Oct 29, 2021 at 8:15
5

You could consider going into AI policy as a field and becoming a government adviser instead of working at a company. What you've studied seems a good fit for such a role. Having said that, you should still try to gain some technical understanding of current AI so you know what it can do and what its limitations are. A good book for that is Melanie Mitchell's Artificial Intelligence: A Guide for Thinking Humans.

For more on AI policy, see https://80000hours.org/topic/priority-paths/ai-policy/.

For a UK example of national AI strategy, see https://www.gov.uk/government/publications/national-ai-strategy/national-ai-strategy-html-version

1
  • 1
    This, very much! The field desperately needs people that can function as a translation layer, between engineers that are too busy to think about the impact of their work, politicians that are behind the times and the general public. A humanities background and profound interest in the field can make you are great asset in this regard.
    – MvZ
    Oct 28, 2021 at 20:20
-9

Two of the most direct ways are:

  1. On the software level, start a commerce business and grow your customers. When your customers reach a certain level such as a billion, you will need AI to optimize the operation. And your competitor is Amazon, Alibaba, etc.

  2. On the hardware level, become a chipmaker. When your chip design reaches a certain level such as 5nm or 3nm. You will have the ability to integrate AI efficiently. And your competitors are Nvidia, AMD, Intel, etc.

These two ways are the fundamental ways to push AI evolution.

For you, you can choose to work for them (become their employee) , work with them (become a contractor), compete against them (work for a smaller, start-up company, or start a similar business of your own).

2
  • 7
    This doesn't answer the question. OP wants to work with AI, not push AI evolution. Also, suggesting to become a billion dollar corporation or a chipmaker isn't particularly helpful.
    – Jeroen
    Oct 28, 2021 at 7:01
  • 1
    Not only this is not an answer, it's also misleading. According to Wikipedia, "Taiwanese chip manufacturer TSMC plans to put a 3 nm semiconductor node into commercial production for 2022,[1] followed by its American counterpart Intel for 2023[2] and South Korean chipmaker Samsung for 2024." Intel is one year behind TSMC.
    – Nobody
    Oct 30, 2021 at 9:28

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .