What is Required to Achieve Career Success?

Non-technical / non-developer roles in AI often don’t get the attention they deserve. In this post, I’ll shed light on why these roles are so critical and how you can break into AI using your current skills and strengths.

There are plenty of resources that guide you in becoming a machine learning or AI engineer or a data scientist. Here I want to draw attention to the many non-technical roles that are essential for AI success, but they often don’t get enough attention.

First, let’s define what I mean by ‘non-technical’. I simply mean, roles in which you don’t have to write or read code, or the work does not require a specialized technical skill such as configuring a server on a cloud platform or applying cybersecurity policies to a database.

Here are 5 reasons why non-developer roles in AI are important

  1. AI is a transformative, general-purpose technology that will impact all industries and domains — your active participation shapes its future.
  2. AI, as an interdisciplinary field, depends and thrives on the collaboration of developers, business experts, engineers, and social scientists.
  3. Creating trustworthy, human-centric, and beneficial AI requires diverse backgrounds and expertise to define values and ethics.
  4. The rise of low-code and no-code tools and platforms has empowered non-developers to create, experiment, and innovate.
  5. With many unsolved problems, AI offers a challenging, fulfilling, and lucrative career for people with many backgrounds.

It’s important to note that a role itself may be technical in the sense that it requires deep domain, engineering, or scientific knowledge but my criteria for this story is that the role does not require AI-specific technical or programming expertise. Remember, there’s not a standard way to work in AI. Even the job descriptions of well-known roles have not been standardized. See below for examples of non-developer AI roles.

The purpose of this illustration is to show that there are many non-technical career areas in AI. A balanced AI team requires technical and many non-technical experts, as shown in the sample list above.

Sample areas for non-technical / non-developers in AI:
1: C-Suite
2: Board Members
3: Finance
4: HR, Enablement/Training
5: Compliance + Risk + Legal
6: Customer Care
7: Marketing
8: Sales

Another way to look at these could be to map them to an AI workflow. This view could be helpful to see how and when these roles come into play during the process of AI solution building.

Why do non-technical/developer AI roles matter?

  • There is a misunderstanding among many students and professionals that AI is mostly about technology. AI is too important to be left to technologists! AI desperately needs people with diverse backgrounds. AI is interdisciplinary by definition and needs a variety of roles
  • The rise of low-code and no-code tools has made it easier for non-developers to test ideas, explore options, and learn about AI, thus democratizing access to AI and paving the way for a new wave of creators
  • As AI is evolving rapidly, new roles are emerging which are a combination of specialized skills with some AI knowledge — whether it’s matching the right data to the use case, prompting language models, validating AI results, looking at ethical aspects of AI, or managing the complexities of AI teamwork
  • By developing your skills, you can build a personal brand and lead a successful and lucrative career!

A Simple Formula to Assess the Value of your AI Role

Here’s a simple way to think about the skill mix.

Value of your role in AI = (A) Domain / Tech Skill Expertise + (B) Fundamental AI/Tech Literacy + (C) People and Soft Skills

For any given role, a person’s technical expertise and domain knowledge could vary. However, there’s a minimum baseline of AI knowledge that a given role needs.

Here’s one way to look at the skill mix for a given role:
AI Product Manager=20% technology/AI + 60% product + 20% industry

Many people are worried about the math or complexity of AI models. This should be the least of your worry! For most of the roles, you don’t need much math or neural network theory. You do need to know the limits of models based on math and statistics and rudimentary concepts of probability. If you don’t understand the basic concepts of AI then it will be hard for you to ask the right questions, validate the assumptions, and contribute at the right depth.

Here’s a mental map for exploring AI careers. Most of the items here are self-explanatory.

Three Areas of AI Learning

A few comments on the venn diagram of skills shown above. Obviously business and other non-technical skills such as strategy, use cases, ethics, compliance and risk, problem framing, design and user experience, cost, management, governance, and all other roles that determine the why and what of AI are always important. There are three specific areas that one should focus on to build your AI foundation:

  1. Data — develop data thinking and learn the modern data stack and lifecycle
  2. Core AI — in addition to the technical concepts and vocabulary, you should also learn why AI is different, how it is shaping the way we do business, and what makes AI successful in different organizations
  3. Technologies that support AI — basic knowledge of the modern technology stack on which AI is built upon

Concluding Thoughts on How to Get into AI as a Non-developer

  1. Build on your strengths: Are you a designer / UX guru, a master at spotting problems, or a data storyteller? Maybe you bring in a unique perspective based on your familiarity with processes or culture. Or you are good at sizing and cost calculations?
  2. Build a solid AI foundation: Learn as much as needed for your role. Combine your strengths with the core data and AI knowledge and skills that you have acquired from a combination of coursework, reading quality articles and papers, and working on real-life (read messy and political) AI problems or projects. Learn how to interpret data exploration charts or how to read the results of models. Create a way to document and record your learning and watch for patterns. Learn tools that make your life easy (for example: ChatGPT and similar tools help those who know how to ask the right question and who learn how to craft the best prompt). More on this below.
  3. Learn with people: No substitute for hands-on learning and group work. If you are embedded in a team, great. If you are trying to break into AI, then network and learn from others. Find projects that imitate real life. Choose learning programs that emphasize the overall challenges, not just the technical aspects.
  4. Common problem: You are given a clean data set, and you start modeling. The key here is to have a growth mindset and become a lifelong learner.