Knowledge

The data and AI literate person knows how to distinguish between different types of data, from traditional structured types such as categorical and quantitative variables, to unstructured data like text, images, audio, and video. Beyond merely identifying data types, they understand what can and can’t be done with them in both traditional analysis and AI applications.

Just as the poet studies vowels and consonants, verbs and nouns, individuals who are highly literate in data and AI embrace the fundamental building blocks of both structured and unstructured data. This enables them to see data for what it is – information that can be grouped, aggregated, summarized, and processed by both human and machine intelligence to generate powerful insights and messages.

The person who is highly literate in data and AI is familiar with ways that data is collected, structured, and stored, and the attributes associated with each approach. They understand traditional storage systems like spreadsheets with their cells in rows and columns, databases with their records arranged in relational tables, as well as modern data architectures that support AI operations, including data lakes and model storage systems.

Similar to how farmers are knowledgeable about growing, harvesting, and storing grain in silos, people who are highly literate in data and AI understand various ways that data is collected, structured, and archived. They’re aware of different ways data is imported, exported, and transformed between various storage types, including how data needs to be organized and versioned for AI model training and deployment.

Those who are highly literate in data and AI understand that storing data is not an end in itself, but rather a means of extracting valuable insight about one’s environment. They grasp the fundamental principles of both traditional statistical analysis and modern AI/ML concepts, understanding when each approach is most appropriate.

Much like an investigator knows what types of deductive and inductive reasoning are useful when solving a crime, a person who is highly literate in data and AI appreciates critical thinking, statistical significance, and what types of insights can be reasonably derived from both traditional analysis and AI-powered approaches. They understand the strengths and limitations of each method, knowing when to apply classical statistics versus when to leverage AI capabilities.

Since, according to Dr. Tamara Munzner, the human visual system is a “very high bandwidth channel to the brain,” the person who is highly literate in data and AI understands various ways to visualize both traditional data analysis and AI model outputs. They grasp the principles of cognition relating to how humans decode visual encodings such as position, length, area, and color, and understand how to effectively represent AI-specific concepts like model uncertainty and decision processes.

Like a doctor who knows how the body will react to prescribed medications and therapies, those who have high levels of data and AI literacy know how their audience members will interpret various visual treatments and forms. They understand how to create visualizations that make both traditional analysis and AI-driven insights clear and accessible to their intended audience.

The person who is highly literate in data and AI understands the landscape of AI systems, their capabilities, and their limitations. They know the major types of AI applications, from language models to computer vision systems, and understand when these tools can reliably augment human decision-making versus when traditional analytical approaches may be more appropriate.

Like a mechanic who knows which tools to use for different repairs, they understand the strengths and limitations of various AI systems. They recognize that AI tools, while powerful, are not magical solutions but rather sophisticated information processing systems that require appropriate data, careful implementation, and human oversight to be effective.