Did you wake up one morning to find out that the whole world was talking about AI and machine learning? From self driving cars to Amazon telling you what you want to buy before you realized you knew you wanted it — AI, specifically machine learning, has crept into every part of our lives.
This brave new world of machine learning (ML) and automation has also sparked a massive debate about the future of work and the role we humans will play in it, particularly for those people (like me) who do not have a techie background.
As an executive at a growing ML company who has a BA in anthropology and urban planning, at least once a week I get asked how can a nontechnical person get their foot-in and ultimately build a career in the emerging AI economy.
Here is the three pieces of advice that I share with folks:
(1) Look for Positions to Train the Machines
Artificial Intelligence is not born intelligent. ML products rely on people to train it on what it is that is supposed to be doing. Think of this as a teacher training children in English language or coach training you how to how deadlift in the gym.
For example, the company I work for, RoadBotics, applies deep learning to help governments have high quality data about their road conditions. Essentially we feed our ML model images of road surfaces and it generates a 1–5 condition rating based on the distresses that are present on the road in each image.
How did our ML model get to the point where it can make accurate and consistent decisions about the condition of any road surface? People had to train it! The rigorous training regimen consists of dozens of people looking at millions of images on a web-based platform and digitally ‘painting’ or labeling the distresses that are present in each image. The resulting ‘labeled images’ are used as the training material for the ML model to continue its learning.
Who are the labelers that are critical to the ongoing training of the AI? People from all sorts of backgrounds — from people with degrees in geology to former Starbucks baristas. In this world of ML training, your academic or employment background matters less because a company like RoadBotics can quickly train you on the objects that you need to label to train the machine.
The moral of the story is that the world of ML can only be created by a critical backbone of people behind AI who are providing continuous training. This presents a major opportunity for those who do not have technical backgrounds to become immersed in some of the most critical work that an ML company has going for it.
Seek out positions that might be described as data labeler, image annotator, or training technician. Although your initial responsibilities will likely be spending hours in front a screen clicking and scrolling, it is one of the quickest ways to get your foot into the door of a machine learning company without a technical background.
(2) Become a technology translator
One of the key reasons that it is beneficial for a non-technical person to seek out a data labeling position is because it gives you a front row seat to better understand the technical world of ML.
Through exposure to the core technology, data labelers can begin to pick up on the technical language of ML, the basic mechanics of the underlying processes of ML and ultimately how companies translate complex technologies (that typically take a PhD or two to understand) and sell products to a lay person like yourself.
And because many of these companies use their labeling staff as a talent pool to promote to other more customer-facing lines of the business, it’s an excellent way to begin to build your career with an ML company with no prior to technical background required.
The key here is to become a technology translator — whether that is through immersion as a data labeler or by seeking out positions that enable you to serve as a communicator between the technical and non-technical.
Often times, this means you will be looking at business development, marketing or customer experience positions. In these roles, you will need to be able to understand enough about the core AI technology to be dangerous, but you do not need to be an expert. In fact, your lack of expertise will help you find new and creative ways to communicate the potential complexity of the tech to an audience that really just wants you to solve their problems.
The faster that you can bridge the technical and non-technical, the faster you become to your team at an ML-centric company. Your engineering and technical team will have confidence in the fact that you actually understand the core technology and can help them grow the business. And your customers that you interface with will have confidence in that you actually know what your are talking about and can help them solve their problems.
(3) Run towards startup opportunities
So, where do you turn to look for positions as a data-labeler or as that tech translator? Without question, look to startups, particularly venture-backed startups.
First, venture backed startups that are building ML products and services require the foundational labeled data you can help create. This typically means that startups are looking to quickly hire the labor force needed to generate a treasure trove of labeled data that will enable them to build, go-to-market and scale faster.
Second, seeking out positions with startups will enable you to get in at the ground level of what could be a successful growth-oriented company. Now, there are massive risks associated with getting involved with a startup and you need to recognize that the majority of startups fail are not the Facebooks of the world. But getting into a startup early quickly puts you in the driver seat to get your foot in the door, become an important member of the team and play a role in building an awesome product!
Finally, the wild west world of startups presents a massive opportunity for accelerated career growth without a technical background. Because startups prioritize speed and growth, you will find an environment where doors rapidly open and new positions get formed that you can be well poised to take. Perhaps most importantly, I’ve found startups to be places where your ideas can matter — a good startup does not care if you are a techie or a non-techie. If you have drive, creativity and ideas that can help the company be successful, you will earn more and more opportunities.
In our AI-driven future, we need more non-techies! I hope these thoughts help more non-technical folks like myself dive into the wild and exciting world of ML!