Artificial Intelligence (AI) is an umbrella name, comprising several different software technologies, including Machine Learning (ML) and Neural Networks. AI Is now operating at a level which can out-perform people at tasks including speech recognition, facial recognition and other ‘human’ tasks that, previously, computers simply would not have been able to complete. More importantly, AI’s ability to undertake these tasks relies not on the programming of specific algorithmic rules, but rather on spotting ‘fuzzy’ patterns (consistent but sometimes vague) and associations which are found in large collections of data.
The field is growing fast. OpenAI, a research lab focussed on accelerating AI and it’s capabilities, has provided some insights into what companies around the world are paying to attract and retain AI professionals. The lab paid more than $1.9m to its top researcher Ilya Sutskever in 2016, and more than $800,000 to another leading researcher, Ian Goodfellow.
In addition to salaries and signing bonuses, many companies typically compensate employees with sizable stock options. AI professionals with little or no industry experience can make between $300,000 and $500,000, per year in some parts of the world, combining their salary, bonus, and stock.
While those mentioned are exceptional candidates and are receiving what any independent observer might describe as sky-high salaries, the fields of AI and ML offer opportunities for anyone prepared to invest in themselves, and willing to learn the necessary skills. Demand for AI specialists has been growing faster than the supply of AI talent and it’s that which is bidding up wages.
Stanford University, for example, produced an interesting reported, sourced with data established through the ‘Indeed’ job search platform. The study revealed that AI and Machine Learning are the most sought-after jobs in the IT space today.
As you see from the chart, the AI job market ecosystem is growing. A new report called “The Future of Jobs 2018.” released by World Economic Forum, (WEF), generated similar results, suggesting machines and algorithms are expected to create 133 million new job roles by 2022. Pursuing a career in AI, could, it seems to make you a millionaire in a few years.
Statistics and coding are necessary skillsets
On the face of it, however, establishing the requisite skills is a difficult task. The problem is the combination of strengths required to meet muster in the field. I’ve spoken to a number of people working in the field of AI recruitment, to establish the core skills required, and where you can get them.
AI is an unusual science because it requires expertise in two very different areas – fields that most developers have little or no experience in. As well as core coding competence, to successfully obtain a position, applicants need to have an intimate understanding of statistics too.
The steps required to build the necessary skills
The dearth of trained individuals with skills in the right areas has created a difficult circumstance. Lured by the high wages, people fake the right skills and attempt to apply for roles when, in reality, they don’t have them, according to those whose job it is to weed the wheat from the chaff (recruiters.)
Success, it seems, having listened to those in the know, comes down to the realization that, while the skills required may be new to you, and while having both an appreciation of statistics AND the more involved aspects of being a developer is a rare combination, both disciplines are well served with training materials. Luckily, today you don’t have to spend years and huge money on University education to become familiar with the AI and ML technology. If you really want to skill yourself up – here’s how you can.
Title: Rate of automation: Human vs. Machine
Enroll with a Free Online Course
A growing number of online courses are available online that you can take without any fee.
Courses are often created by some of the top universities in the world and cover everything from the basics to advanced implementations of this seemingly complex technology. The very fact that some of the biggest names in technology are prepared to offer these materials free to the community talks to the gap in the labour market, which needs to be filled.
For example, Stanford University’s free ‘Introduction to AI’ is, many say, the best place to start learning AI. Most of these classes can be undertaken anywhere, on the bus to work, on the weekend, from an internet café – since they’re all available through your mobile phone’s data connection. There really is no excuse!
Microsoft, too, offers plenty of free AI related training material online. The company provides two tiers of courses – one aimed at ‘Beginners’ and, another targeted towards ‘Advanced learners’. They’re free and, importantly, when you successfully complete them, you’ll be adding an endorsement to your CV, from a brand which potential employers will recognize and esteem.
Finally, Google’s TensorFlow is the most popular open source AI software library. It features a number of existing code depositories and allows you to experiment with neural networks in various ways. Google’s free courses take a total of around 2 full working days, (15 hours) and consist of a total of 25 lessons and 40 exercises to ensure you are grasping the concepts they’re laying down.
Take Paid Courses – MOOCs
Alternatively, if you have deep pockets, you can choose from paid Massively Open Online Courses (MOOCs) structured and delivered by Universities’ professors or expert trainers for a fee.
Some of the best websites offering paid online courses in AI and machine learning are Coursera, Udemy, CodeAcademy, Team Treehouse, and Lynda.
These websites come with a curriculum created by some of the real, found experts in the field (for example, Andrew Ng.). Most of these platforms come with interactive video lectures and downloadable resources.
The nature of this sort of learning means that you can work through them at your own pace and within your personal schedule. In addition to MOOCs, there are a variety of relevant books available, again, featuring a lot of expert material, tutorials, and examples, to make learning more interesting and to help you retain what you’ve experienced.
Do something special – extra credit to set yourself apart from other applicants
Two things came out regularly when discussing this subject with recruiters. Applicants for jobs in the field enjoyed more success when they had something special about them. Here are 2 things you could work on.
One positively differentiating factor is having a patient to your name. Patenting something may seem daunting if you’ve never done it before but the process is designed to be followed by those with little experience and shows that you are on the cutting edge of your field.
Finally, and perhaps, for some, more achievable, entering a competition is a way to show passion (again), learn new skills, network and, potentially, to stand out from the crowd for all the right reasons. One good example, of a brand of competitions to enter, is Kaggle, (see screenshot below) a community for Data Scientists. In it, they share code and compete to find better ways to address some of the challenges AI provides.
There’s a burning platform as well as a $1m reward
Simply the act of making an investment in yourself, in terms of your time, money and passion for the area, shows potential employers just how serious you are about AI, and could open doors to better opportunities and higher salaries.
If the lure of high wages is not enough to entice you, ask yourself which side of the future you want to be on. As you ‘ve seen from the above, soon, many human roles will be displaced and replaced with code. Building the AI software which does that might be one of the few positions left.
Make no mistake. None of this is easy. Simply because many of the resources are free to use online does not change the fact that to skill up will require months of out of hours, personally motivated, self-development. If you work at a company prepared to sponsor you in some regard (most usefully, by exposing you to projects which allow you to ‘play’ safely with these new technologies), the task may be that little bit easier.
As the world is rapidly accelerating towards the widespread adoption of AI technologies, the job industry will adapt to the opportunity. It’s worth considering the issue from a higher vantage point, however. The government should be considering how it can use educational institutions to most effectively teach the people who are displaced, the skills they need to keep up with the modern economy.