On what resources do management strategists need to focus, if they want to leverage Artificial Intelligence and create Competitive Advantage? To provide an answer from business academia, a framework for leveraging Artificial Intelligence in a business context is presented in this 5-part article series. In Part 4, the relevance of skilled labor is elaborated, especially for the role of the Data Scientist.

AI is changing our work environments and the needs of employers

The third type of resources you need for applying AI in a business context are AI-enabling human capital resources, concretely analytically, managerially and technically skilled labor. The rise of AI is resulting in a so-called “Skill-Biased Technical Change” (SBTC), defined by Bresnahan as a “technical progress that shifts demand toward more highly skilled workers relative to the less skilled”. He states three arguments why AI or IT in general causes SBTC:

  1. Complementation Effect: IT-related skilled labor is complementary to IT infrastructure, because IT infrastructure always requires people for management, operation, maintenance, etc. Since IT hardware prices are declining rapidly, the demand for complementary goods are rising by basic economic principles.
  2. Substitution Effect: IT, and especially AI, is often used to automate routines and well-defined tasks, therefore substituting clerical or blue-collar workforce. On the other hand, more complex work (like those of managers and professionals) is hard to substitute, so there is only “limited substitution”.
  3. Information Overload Effect: There is an effect of information overload, meaning that computerized business processes accelerate data generation, which in turn causes more demand for skilled labor.

High Demand + Low Supply = High Price

This high demand for skilled workers caused by SBTC contrasts the shortage of supply in the labor market. While the most immediate need remains critical technical skills, companies are also encountering a great lack of people at the interface between business and AI as well as system and data engineers.

A study by McKinsey&Company differentiates between three key types of skilled workers:

  • deep analytical talents for statistical modeling and knowledge discovery,
  • data-savvy managers and analysts for interpretation and decision-making, and
  • supporting technology personnel for maintenance and data processing.

They forecasted that demand for deep analytical talents in the USA would exceed the projected supply by 50% to 60%, as well as a need for 1.5 million additional data-savvy managers and analysts. This disequilibrium between demand and supply of skilled labor results in enormously high salary offers by tech companies. The New York Times reported that typical AI experts (including PhDs and less educated people with work experience) can earn between $300,000 to $500,000 a year in salary and equity.

The "Sexiest Job of the 21st Century" – or not?

This general role of the analytical specialist is widely known as “Data Scientist”. Even though it has been called the “sexiest job of the 21st century”, there is no precise definition of what a Data Scientist is, or what skills this role comprises. Many definitions from business as well as from academia describe an all-round talent, incorporating knowledge in Data Science, statistics, Machine Learning, programming, data visualization, the specific business domain, and at the same time being a team player and excellent communicator. However, to put it in a nutshell:

“Data Scientists discover new patterns in data, using sophisticated statistical methods from Data Mining or Machine Learning, to realize business value by supporting or automating decision-making”.

Thus, the discovery of new, business-relevant knowledge from data is the key purpose of Data Scientists. This is especially hard in a business context, since real world problems are rarely aligned for applying AI techniques directly. The Data Scientist needs time to exploratively experiment with the data and the freedom to build and test hypotheses, prototypes or proofs-of-concept.

Unfortunately, this challenging, creative work is often overshadowed by rather simple, time-consuming data processing work. The reason for this is that the role of the Data Scientist is still not well understood. Like chemistry in the mid-19th century, the field of Data Science is not yet well established, and a good Data Scientist must not only be proficient on the scientific level, but also be a good “lab technician”, doing most of the preparatory work himself. This lack of understanding might be one of the key issues in hindering organizations from realizing the full potential of (big) data. To tackle this issue, business and academia should collaboratively define the required knowledge, roles and skill sets across the organization.

Skilled Labor as a Source of Competitive Advantage?

To summarize, since analytically, managerially as well as technically skilled-labor is very rare, but a necessary complement to make use of AI at the same time, it indeed can be a source of Competitive Advantage. This point stands in contrast to Mata, who argue that technical skills cannot be a source of Sustainable Competitive Advantage, since they are not heterogeneously distributed and highly mobile. On the other hand, they point out that managerial IT skills can be a source of Sustainable Competitive Advantage, due to their long-term development and social complexity.

However, more than 20 years later, it is debatable if this argument still holds for AI-related skills, since they actually are very rare and heterogeneously distributed. Anyway, the good news is that AI-skills are spreading quickly, thanks to the increasing amount of online educational resources as well as novel curricula at universities. But for the time being, managers need to balance out the benefits of gaining more business value from their data, against the high cost of skilled labor.

In the last part of this 5-part series, the relevance of AI Knowledge is elaborated, especially the role of open source Machine Learning algorithms. Stay tuned!

Show Sources
Unser Newsletter

Abonniere und verpasse keine Neuigkeiten

Vielen Dank! Jetzt bleibst du immer auf dem neusten Stand!