Where Do World Leading Companies Get Their AI Expertise From?

July 20, 2018


facebook icon facebook icon

For many years, the main goal of companies is to collect as much user data as possible. Dealing with all these and new incoming data quickly and effectively is impossible without intelligent systems. This is why companies desperately need to harness AI technologies to come to the top place among competitors – and the sooner the better.

However, the challenge is that modern AI systems are “idiot savants” as Gurdeep Singh Pall of Microsoft put it in one of his talks. “They are great at what they do, but if you don’t use them correctly, it’s a disaster.” This means that the technologies are still weak and highly depend on human genius.

The biggest obstacle is that the people who have expertise in AI are rare. As stated by The New York Times, today there are fewer than 10,000 specialists who have enough competence and expertise to solve serious AI challenges.

When fighting to hire an AI specialist, companies do not disdain any methods, sometimes acting aggressively and overbidding each other by proposed salary sums.

According to 2018 Glassdoor research, Facebook, NVIDIA, Adobe, Microsoft, Uber, and Accenture are among the most attractive companies searching for AI experts. All together, they are currently hiring for 96 open AI positions. Let’s consider the most popular techniques of acquiring AI talents used by the world’s leaders. First, we suggest you look through the infographic combining the most interesting facts on how leading companies make AI expertise for their peer-to-peer marketplace websites:

Academic Brain Drain

The main source of AI expertise for leading businesses is technical universities and academies. Major companies literary bite the best pieces out of the academic staff roster and the universities are then helpless to it, unable to compete with salary, computing facilities and technical challenges proposed by the tech labs of world-wide known IT empires. Yesterday’s university professors have become leading experts in Google, Facebook, Uber, Amazon and other giants with six-figure salaries. There is no rare case when promising undergraduates leave their studies to join the Googles or Apples of the world.

Tech conferences, meetups and private parties have become easy platforms for talent shopping.

Thus, Sheffield University lost its professor of machine learning, Neil Lawrence, who joined Amazon researchings. Cambridge said farewell to Zoubin Ghahramani, their head of machine learning who set out for the position of Chief Scientist at Uber. Imperial College London said goodbye to Murray Shanahan, a Professor of Cognitive Robotics who, together with Oxford’s professors Yee Whye Teh and Andrew Zisserman, was hired by Google.

It became a real sensation when, in 2015, Uber managed to grab 40 researchers at once from the robotics faculty at Carnegie Mellon University. Later, to compensate the loss, Uber paid $5.5 million to the university department.

The best AI brains are concentrated in several leading companies and this emerges an even bigger shortage of scientists. Without good university professors, it becomes difficult to educate a new generation of AI specialists while the demand for them only grows. “It’s like killing the geese that lay the golden eggs.” sais Zoubin Ghahramani, professor of information engineering at Cambridge University and Chief Scientist at Uber. “Companies are starting to realize that and some of the major tech companies are starting to give back to universities by sponsoring lectureships and donating funds.” Many professors continue to combine researching work for companies with teaching practice in Universities or within companies’ inner educational departments.


One more commonplace practice used by companies when there is no other opportunity to get the necessary specialist or technology is a so-called acqui-hiring (a term born from blending ‘acquiring’ and ‘hiring’). Companies use “what’s yours is mine” approach when acquiring the whole startup that possesses valuable resources like employees and knowledge base.

As reported by RS Components, over the last two decades, the largest acquisitions in AI sphere were made by:

  • Google – $3.7 billion for 29 AI companies
  • Amazon – $821 million for eight AI companies
  • Intel – $776 million for five AI companies

It’s not only IT-specialized companies that invest in Artificial Intelligence technologies. For instance, in 2017, Ford Motor Company acquired Argo AI for $1 billion. In the moment it was acquired, Argo was a six-month old startup, founded by Bryan Salesky and Peter Rander, Uber’s former engineering lead who contributed to the first-generation self-driving prototypes.

A sports clothing brand, Nike, has also recently bought an Israeli AI startup, Invertex, to improve the customer experience providing foot-shape recognition.

Thus, step by step, AI technologies penetrate into each sphere of business.

Raising your Own AI ‘Army’

Many companies do not want to waste time waiting years while universities bring them fresh graduates. Instead, they realize that it makes sense to democratize AI technologies for their onboarded employees who are already great specialists and just lack AI knowledge. For this purpose, such companies as Google, Facebook, Uber, and Airbnb have created classes and research internships to teach the engineers who lack the background in deep learning and neural networks. “We have incredibly smart people here,” says Larry Zitnick, a lead at Facebook AI Research. “They just need the tools.”

Airbnb, a leading vacation real estate rental service, has already been actively implementing the power of AI into their predictive mechanisms to enhance customer experience with pricing and personalized suggestions. As they need more and more connections between the Airbnb platform with ML, they need experts who are able to build it. Instead of hiring new employees with AI knowledge, they decided to educate their own team. To do this, they founded a Data University where any employee can attend classes to gain knowledge.

Another approach was selected by Uber that pays great attention to ML to maximize the accuracy of their predictive models. To eliminate the necessity of creating bespoke ML models for each new project, Uber created Michelangelo, the ML-as-a-service system, that allows users to quickly build, deploy and operate machine learning solutions by all employees from any department.

Teaching AI to Create Other AIs

As everyone understands, AI is our future that will substitute old approaches and change the perception of many of today’s business processes, and huge companies are already engaged themselves by creating methods to facilitate AI growth. Consolidating the best minds, Google, Microsoft, and other “big guys” have already started working on a creation of AI that is capable of creating their own AIs, thus eliminating the need to hire more human experts. The goal is to help democratize AI for small and mid-sized businesses that cannot afford their own AI researchers and labs.

Satya Nadella, Microsoft’s Chief Executive Officer, is confident of the importance of making AI available to everyone, especially in spheres such as healthcare, education, manufacturing and retail: “We need a technological breakthrough that drives growth beyond us.”

Bottom Line

To sum it up, it is clear that AI is no longer science fiction. It is our reality that is going to revolutionize the whole business world in the very near future.

With the deficiency of AI specialists, there are several ways to build AI expertise in a big company. The most prominent university students, graduates and professors become targets for headhunters. Hiring rare specialists, educating own employees, acquiring AI startups and creating ML-as-a-service systems are the main methods that allow companies to get artificial intelligence expertise.

While the giants of IT fight with each other for leadership in the AI domain, smaller businesses can sit back and relax, waiting for the development of massively available AI-as-a-service solutions that will allow them to build ML models for their needs.


facebook icon facebook icon

Anna Klimenko

Anna Klimenko, a technical writer at Greenice. Anna is passionate about technologies and innovations. Writing for the company’s blog, her aim is to translate the technical language into human language and give examples of how you can apply the technologies to your business.


Sociable's Podcast