AI: the disruptive force that will transform L&D
According to Bloomberg, AI is likely to be the most disruptive power in technology in the next decade, which will be critical to learning and development in global organisations.
Two of the most critical difficulties in L&D today are made by the widespread lack of valuable metrics and data-driven industry insights:
- Employee Skills Gap Analysis.
- Calculating ROI of Learning Applications.
Many businesses still train for skills that don’t meet their actual requirements because they fail to manage and understand the right data to recognize skill gaps and drive market outcomes.
AI-powered changes by companies such as IBM, who can now divine future skill gaps and performance, are pointing the way to solve these difficulties.
By leveraging Artificial Intelligence to identify improvement needs and link learning-related metrics to business results, companies can create compelling, self-optimizing training programs. AI-powered programs can make employee-specific forecasts and support for skill development, future production, and collaborative training experiences by ingesting and investigating various data sources.
Most businesses are already applying Artificial Intelligence to analyse large volumes of unstructured data to draw company conclusions from, which is a sign that this is a logical next move for L&D and HR in general.
Description of AI and how it predicts the future
In market and technology, the term Artificial Intelligence is often used as a sunshade term to refer to independent machines, software, and algorithms with training or problem-solving capabilities. This combines the ability for machines to understand natural language to interact with humans, make forecasts based on data to drive penetrations, analyze physical surroundings to push a vehicle, stimulate the brain's neural networks to identify images or translate text, and so forth.
One major subfield of Artificial Intelligence is machine learning. There’s no agreement on the exact definition of machine learning, but the following description on Nvidia’s blog describes the inner workings of machine learning quite well:
“Machine Learning at its most basic is the practice of using algorithms to parse data, learn from it, and then make a determination or prediction about something in the world”.
A well-known subset of ML is deep learning, a technical idea based on artificial neural systems that makes computers able to learn automatically without entering hand-coded rules.
ML and deep learning algorithms are generally available and accessible via open-source software libraries and cloud computing applications such as IBM Watson, Amazon Web Services, Google Cloud, and Microsoft Azure. Essentially, this means that anyone with the right set of data can make AI-powered forecasts.
A quote by Ted Sergott, EVP, Product Development at PRO Unlimited and part of the Forbes Technology Council, decreases the strength of data-driven AI as:
“One of AI’s biggest strengths lies in its ability to sift through millions of unstructured data points and make sense of them quickly.”
In industry and HR, platforms like IBM Watson mix deep learning technologies with other Artificial Intelligence technologies and complex business-oriented data sets to make forecasts on employees and business achievement. IBM Watson can now, with 96% efficiency, infer the possible skills and qualities workers might have to serve IBM in the future by looking at an employee’s skills and projects. IBM administrators take Watson’s likely future performance assessment rating into account as they make reward, pay, and promotion choices.
Bloomberg records: “To motivate employees to learn new skills, more employers are beginning to focus on the future through evaluations. The shift to skills-based production management is motivated, in part, by employers that say they’re fighting with a skills gap.”
The potential of Artificial Intelligence to help employees develop the right talents and to help employers predict future requirements is massive.
The problem of identifying skill gaps
Historically, an essential reason that education programs haven’t yielded the wanted results is that they’ve usually been launched without sufficient information of where the gaps in employee skills exist. Millions of dollars have been funded in learning programs without understanding the need and impact of these plans. With the rise of social enterprise, this dilemma has become even more difficult.
Recent study by Deloitte Human Capital Trends, in a statement, titled The rise of the social enterprise, concludes6 that in the 21st century, careers are no longer narrowly defined by jobs and skills, but through experiences and learning agility. But, only 37% of the 11,000 business and HR manager respondents believe their companies are ready for this shift.
In fact, 72% of companies already have job paths that do not follow traditional organisational authorities, while only 53% of companies develop their people through collaborative education and experiences needed to develop talent following these nontraditional career paths.
According to McKinsey & Company, unleashing the potential of collective intelligence is important to the digital-learning change. Employees learn through collaboration and can be allowed to share information across the company, which can be used to identify skill gaps and measure achievement.
In big global companies, HR or L&D can’t own specific knowledge about the existing and emerging experiences a diverse workforce must have to improve the appearance of each business. In these situations, the collective wisdom of all employees can be used to recognise existing and emerging talents. This necessitates the collection and analysis of data from employees on an extensive range.
Consequently, to identify skill gaps in the current workforce, companies would benefit from a holistic way to data collection and intelligence.
Learning-related metrics and business value
The absence of learning-related metrics and complexity ensuring the continuous development of skills are among the most pressing difficulties in their skill-building programs, officials report in a survey administered by McKinsey & Company8. Metrics are a necessity for building skills sustainably. More than half of managers say that their organisations link employees' skills in learning applications with individual performance. Still, only 12 percent say their companies calculate the quantifiable returns on their training investments.
The four levels of the Kirkpatrick Model, a scientific basis for estimating the effectiveness of learning answers, all need data collection to conclude from. A significant challenge facing L&D is to be ready to make sense of the data and to leverage those insights to drive business value.
Analysis by LinkedIn Learning means that the top three signs of success of L&D programs, according to star developers, people managers, and executives, are an improvement in performance metrics, retention, and qualitative behavior change.
Moreover, McKinsey’s research further shows that organisations that build skills most definitely do a better job of connecting skills to performance and targets than businesses that don’t build skills as effectively.
This suggests the question: how can companies effectively measure the effectiveness of learning applications and link these metrics to market performance to understand ROI?
Predicting skill gaps and measuring ROI
The difficulties of identifying skill gaps and building effective composite metrics are closely associated. They both need an ongoing loop of data collection, analysis, and optimisation. The issue of what skills to develop can be answered by analysing people data and market outcomes. AI can be applied to automate this loop and self-optimize the method based on machine learning principles.
By investigating an employee’s experiences, plans, and training programs, machine learning models can divine what skills the worker should work on to best serve the business in the future. This is what IBM Watson Analytics is now doing to predict employees' skill growth needs and future performance. Watson Analytics is a program for business analytics and intelligence (BI), which uses predictive ML models and cognitive skills to draw business results from data.
Platforms like Watson can examine and combine large volumes of structured and disorganized data to draw industry insights from. This means that rather of having to feed the program data in a particular format, the program can ingest anything like xAPI records, databases, csv files, a stream of events, document templates like surveys, and emails. Artificial Intelligence models have been trained to know all these data inputs, making them able to interpret and structure the data automatically.
These AI-driven BI programs require businesses to only provide data pipelines from their current systems, like HRISs, LXPs, LMSs, and survey tools. From these data pipelines, dashboards displaying metrics such as current skill gaps, future skill gaps, ROI, and worker satisfaction can be built. Moreover, the AI-enhanced data can be pulled back into training platforms to enhance and optimize learning activities automatically.
A great advantage of using a business intelligence program is the ability to involve domain specialists and analysts in creating the relevant dashboards and prediction rules.
Instead of manually programming any dashboards and reports, companies can use intelligent AI to automatically learn from data while using non-technical domain specialists to convert these trainings into valuable business shrewdness. This is now a common tradition in many industries like healthcare, where doctors and investigators work together to build useful BI dashboards without requiring programmers to build these dashboards. Hence, this seems like a reasonable next step for HR.
This ability to make knowledge of large-scale, sparse data is why AI is instrumental to getting skill development needs and linking these to business results. Knowing the relations between many data sources leads to the best understanding of worker performance, skill possession, business administration relationship, and ROI. A human being or a general software application would not be able to do any of this.
Facilitating learning experiences with AI
The organisational development from classical, hierarchical structures to social happenings has contributed to the need for collaborative, human-centric learning activities. In the future, Artificial Intelligence could help facilitate these events by connecting people, suggesting sources, learning automatically, and optimizing equally.
Artificial Intelligence can use people's data, surveys, feedback, and content characteristics to match people to other people and resources to construct essential learning skills. By analysing actual matches and continuous learner feedback, ML models can automatically suggest learning activities and self-optimize. This way, people can be matched with the best guide and resources to support their mentorship program, or with the right peers, specialists, and resources to collectively go through a significant group learning experience.
In addition to data-focused Artificial Intelligence running in the background, AI can directly improve the user experience. By knowing what learners have been working on and dynamically recommending relevant links and activities, Artificial Intelligence can help the user naturally learn and grow. Like conversational AI and chatbots, other AI technology can fasten at key points in the user journey to provide proactive direction and more effectively collect context-aware feedback at natural moments in time.
Combining machine learning AI, conversational AI, agent feedback, learning data, and outside data sources could lead to a program capable of automatically promoting learning activities. As Artificial Intelligence learns more about the learner, it will get better at recommending content, people, and learning pathways. This way, students can follow customised learning activities, automatically optimised to fit their wants and needs.
#GlobalAIandDataScience#GlobalDataScience