Dissecting AI Skills Through Time – Explore Interactive Data
In today’s rapidly evolving job market, staying relevant is no longer about mastering a single skill or sticking to long-standing roles—it’s about adaptability and continuous learning. This is especially true in the Information Technology sector, where advancements are transforming the nature of work at an unprecedented pace. As artificial intelligence and digital technologies continue to reshape industries, many traditional skills are becoming obsolete, while new, high-demand skills are emerging at a fast pace.
Focusing on emerging skills, rather than holding on to those that are fading, is crucial for both individuals and businesses. Workers who acquire in-demand skills such as machine learning, data engineering, cloud computing, and AI algorithms are better positioned to seize new opportunities, while companies that invest in upskilling their workforce can maintain a competitive edge. In contrast, neglecting this shift can lead to widening skill gaps, underemployment, and a workforce that struggles to meet future demands.
In this article we give a glance of emerging versus disappearing skills in the Information Technology industry and how such skills are disrupting the industry at a rapid pace.
When discussing emerging versus disappearing skills, the key lies in tracking how their presence in job postings evolves over time. By examining the frequency of specific skills in relation to the total number of job listings across different periods, we can identify skills in demand and quantify their relative importance. The following word clouds give a glance to the legacy versus in-demand skills in the IT sector, weighted by ranked scores developed by our team.
For instance, in 2010, skills such as CHAID, ROLAP, Pipeline Pilot, and IDEFIX were relatively more prominent. In contrast, by 2021, demand had shifted toward skills like Artificial Intelligence, Machine Learning, and Deep Learning, which appeared more frequently in job postings.
Skills should not be viewed in isolation. To remain competitive in a rapidly changing job market, individuals must strategically focus on building a portfolio that integrates emerging, high-demand skills while transitioning away from legacy skills. This approach not only boosts employability but also ensures long-term relevance in a dynamic employment landscape.
AI is not a singular entity, but a broad field with multiple interconnected disciplines and application areas. Grouping related skills helps illustrate the diversity within AI and analyze development more systematically. This approach highlights the breadth of AI applications, showcasing how fields like data science, machine learning, generative AI, etc, contribute to the larger AI ecosystem.
Grouping skills also helps us track how specific areas of expertise have evolved over time. The job market is dynamic, and demand for skills often fluctuates as new technologies emerge and older ones become obsolete. We have defined 7 groups : Data Science, Data Engineering, Machine Learning, Deep learning, General AI skills, Speech Recognition and Generative AI. For example : ‘Data Engineering’ skill group consists of skills like ‘AWS Elastic MapReduce’, ‘Databricks’, ‘PySpark’, etc.
It is important to note that a skill may belong to multiple groups based on usage and methodology, but has been assigned to a single group for our preliminary analysis.
The previous figure illustrates the number of jobs requiring specific skill groups in different years. Data Engineering has been the most in-demand skill group from 2010 through 2023. This is followed by Data Science until 2017, which is taken over by Machine learning 2018 onwards. Deep Learning and General AI skills see similar increasing and decreasing trends over time. Notably, only Generative AI experienced a slight increase in demand in 2023, while demand for all other groups declined.
To gain deeper insights, percentage changes in skill demand for each cluster were plotted. Percentage changes highlight the relative growth or decline of skill demand over time, allowing for meaningful comparisons across clusters with different starting values.
In 2023, the demand for Generative AI skills surged by an astounding 2,386.7%, a dramatic increase that eclipsed all prior growth in the field. This unprecedented spike was driven by the rapid and widespread adoption of advanced technologies like ChatGPT (released in November 2022), which reshaped industries and created a flood of opportunities. Prior to this monumental rise, Generative AI had experienced volatile swings, with a 300% surge in 2012 followed by declines in 2013 and 2015. Although the percentage increase is striking, the actual number of jobs requiring Generative AI skills grew from 30 to 750, which is still a relatively small base compared to more established fields. It is also important to note that Generative AI is the only skill cluster that saw a positive percentage change in 2023, while all other skill clusters experienced a decline. This sharp contrast underscores the growing demand for expertise in this emerging field, driven by the rapid advancements in AI technologies. As the technology matures, we can expect further significant increases in both the number and diversity of roles that require Generative AI expertise.
The following figure removes Generative AI from the analysis, allowing us to observe patterns in other AI-related roles without the overwhelming influence of the recent spike.
The rise and fall of skill demands are a natural course of a technology adoption cycle. Initial surges occur due to innovation and integration into mainstream applications, driving the demand for specialised expertise. As the technology matures, demand stabilizes or starts declining due to market saturation or shifts towards broader or other emerging technologies.The adoption of technologies is often marked by key developmental milestones, such as the release of voice recognition tools like Siri (2011), breakthrough neural networks like AlexNet (2012), frameworks like TensorFlow (2015), and advanced models like GPT-3.5 (2022).
Speech recognition skills experienced a dramatic 170% surge in demand between 2015 and 2017, reflecting significant advancements and integration into mainstream technologies. However, this was followed by a sharp decline of approximately 225% by 2019, highlighting a rapid demand reversal within just two years. Similarly, deep learning experienced some of the most significant percentage increases in demand, particularly in 2012 and 2017. While most of these skill groups saw an overall decline in demand compared to the previous year in 2019, they quickly picked up in 2020 to only plummet again in 2023.
Overall, 2023 stands out as a challenging year for most skill groups, marked by collective broad declines, reflecting a possible reduction in job postings or a redistribution of focus toward emerging technologies. These patterns highlight the evolving dynamics of skill demand in AI, driven by current requirements.
These milestones are not presented as causal factors for shifts in skill demand but rather as indicative markers of technological progression. The purpose is to contextualize the evolution of AI technology and how they often coincide with markers of technological innovation