Data analysts use artificial intelligence (AI) for a variety of tasks, such as data cleaning and visualization, to reduce the time to insights. Currently, AI-powered solutions are helping analysts forecast trends, perform predictive analytics, and even tell stories through auto-generated summaries.
This is essential for efficiency in the modern workplace, where bulk data from diverse sources pours in from many directions, and speed is of the essence. However, it does offload key cognitive functions to these advanced machine learning (ML) algorithms and large language models (LLMs).
For example, consider an e-commerce store. AI can quickly spot unusual drops in sales across regions and provide insights on why. Human analysts will manually need to check sales data, compare it with past records, and drill deep into traffic or engagement metrics, which is time-consuming.
Tasks that require critical thinking, like pattern recognition, comparative analysis, and diagnostic analysis, are now conveniently offloaded to generative AI models. Gradually, this can erode the central skills around data analytics, such as formulating hypotheses, questioning data sources, or looking at the same data from a different perspective.
Just like unused muscles gradually atrophy over time, cognitive abilities also take a hit when these functions are outsourced on a consistent basis.
Consequently, decision-makers will start taking AI’s suggestions at face value, which can backfire because machine-generated analytics lack depth and reason. It lacks human context, intuition, and domain expertise, leading to recommendations that may ignore practical viability, ethical concerns, or long-term business impact.
As AI grows in capabilities and becomes more accessible to data teams across businesses, teams need to recognize the trade-off between speed and depth to mitigate the problem.
In this article, let’s look at how AI adoption affects the intellectual rigor of human analysts and what can be done to prevent it.
How AI Helps Data Analysts, and Where It Falls Short
AI facilitates business planning for critical functions, such as staffing, budgeting, inventory, and marketing. The algorithms forecast demand, simulate workflows to identify risks, and optimize resource allocation to enhance operational strategy.
These solutions give recommendations based on user actions and market conditions. They suggest data-backed steps to take next. These tips can help fast-moving teams remain agile.
Fundamentally, data analysts use AI to narrow their focus to the data points and insights that matter. It quickly sifts through vast datasets to create meaningful data visualizations or summaries that direct action.
This is great for line-of-business citizen analysts, where professionals usually have to make straightforward decisions, such as restocking inventory or hiring more staff.
The datasets to be considered are structured and repetitive, making it easier for deep learning AI to draw actionable conclusions. Moreover, the outcomes of these decisions can be easily measured and predicted, even with intuition, reducing risk further.
However, for nuanced decisions, such as which product feature should be prioritized and how to charge the clients, data analysts need more. These decisions involve a bit of ambiguity.
There are usually tradeoffs and contexts that can’t be captured by raw data. For instance, feature A may satisfy more users, but feature B might increase revenue and market authority in the long run.
The CTO and co-founder of Pyramid Analytics, Avi Perez, explains: “Usually, LLMs are good at providing qualitative responses, and they excel at interpreting natural language and mimicking humans in their answers. But the downside is that they are lousy at computing real-world mathematics or performing analytic processing. In fact, they are most likely to give inaccurate responses, if not the wrong answers.”
Here, it’s pivotal to provide human judgment, cross-functional input, and emotional logic.
Interpreting unstructured or incomplete data, storytelling for streamlined stakeholder communication, and critical thinking to avoid biased decision-making remain uniquely human strengths.
And the associated cognitive skills of the above strengths can erode away when data analysts constantly outsource them to AI applications at various scales.
Are Human Analysts Losing Their Edge?
Earlier, people navigated with maps. However, as GPS technology made its way into smartphones, map reading became a rare skill. While navigation became simpler, GPS reduced spatial awareness and mental mapping. And these skills are useful in data analysis as well.
Spatial awareness helps decision-makers spot outliers in datasets and visualizations, such as heatmaps. Mental mapping connects measurable efforts with tangible results, aiding operational alignment across the organization.
If humanity has somewhat lost valuable skills due to the shift from paper maps to GPS, one can only imagine the impact of AI on data analysis.
Conversational AI chatbots can ingest real-time data to generate visualizations and offer insights. This ease of use can encourage human analysts, albeit gradually, to rely more on AI systems for data-driven decision-making. Subconsciously, when determining the future course of action, business professionals will refrain from out-of-the-box thinking, which is essential for innovation.
A recent study conducted by MIT confirms this as well. It compared the brain activity of two groups of people: one used AI to write essays, and the other didn’t. The AI users demonstrated reduced internal neural connectivity and memory retention compared to those working unaided.
Tech journalist Gina Marrs summarized this phenomenon: “The ease and speed they [AI bots] offer may actually discourage deep thinking, critical analysis, and the effort needed to fully understand complex ideas. Instead of struggling through a problem or forming original arguments, many users now lean on AI to generate answers instantly.”
Organizations and data analysts need to realize the long-term price of cognitive erosion in the exchange of fast analytics: diminished analytical rigor. The analysts may become less likely to challenge AI’s suggestions and reduce the time spent thinking.
An effective way to retain or even enhance the foundational skills is to revisit them regularly. While it may seem like a chore, rehearsing skills such as manual model-building, hypothesis formulation, and statistical reasoning can enhance cognitive abilities and mental endurance.
Furthermore, whenever possible, it’s generally a good idea to discuss AI’s outputs with other team members to ensure its potential efficacy. Perform cross-checks and manual explorations to validate AI’s suggestions before implementing them.
What Data Analysts Must Embrace
It is clear that AI-driven platforms will be used for data prep, synthetic data generation, visualization, and analytics. Data analysts should, therefore, master the AI tools used in the process. Teams should invest in learning about the limitations of solutions when extracting actionable insights from raw data.
This is crucial for querying the data correctly. Professionals, once they realize what AI-powered analytics software can do, will use it to complement their abilities, enhancing the overall process.
Additionally, it will be easier for data analysts and decision-makers to spot hallucinations and other inconsistencies in AI’s suggestions. Organizations can assess the trustworthiness of models and can use them effectively for various analytics use cases.
Active vigilance, by keeping humans in the loop, toward responsible AI models, will ensure ethical oversight, bias detection, and effective data governance. Analysts can leverage their experience and domain knowledge to validate AI insights and implement them carefully.
Julius AI’s Connor Martin highlights: “Ethical considerations naturally come to the forefront. Some of those ethical concerns relate to the impact AI technology has on society and an organization’s cultural values.”
Adding humans to AI data analysis helps with evaluation. Teams can tangibly determine if they are using AI-powered tools effectively to support their decision-making.
Wrapping Up
AI is now a central part of a data analyst’s toolkit. Various action items, such as data prep, visualization, and summarization, are now expedited with AI.
The growing role of these advanced solutions can lead to cognitive debt, where human analysts lose their critical thinking abilities when deriving insights from data.
Therefore, it’s pivotal to understand how AI works and use it at appropriate moments. The broader objective should be to enhance analytics with AI rather than replacing humans.
A balance between speed and depth is pivotal for the ethical implementation of AI in analytics workflows of businesses, ensuring data security and governance.
