# AI Generated Data Visualizations By:: [[Brian Heath]] 2023-05-08 Today AI chatbots can reasonably respond to natural language questions with natural language answers. Sometimes these answers are accurate, and sometimes they are not accurate at all. However, the responses are good enough that they feel and look authentic and not robotic. Language models have improved significantly in the last few years. This is due to the massive amounts of written content on the internet. There is plenty of context and content for it to learn from. But what about all of the other data on the internet that is not in natural language format? For example, what about all the primarily numeric data stored in relational databases? There is often little to no context for this data unless someone took the time to write down the definitions of each data element. Here AI continues to struggle because numbers largely lack all context. This is why there are so many analysts out there. Many people view our role as strictly making sense of these numbers. While this is one part of the job, it is hardly the only thing. Yet, many analysts only live in this space. These analysts should be worried as it is only a matter of time until AI takes this job as well. Creating good and interesting visuals from a dataset is easily accomplished via AI methods. Many companies and AI bots are already well down this path, and it's only a matter of time. If your job is only to generate interesting visuals, I highly recommend you expand your definition of analytics. Remember, these visuals still have no context. Provide the thought, meaning, and future implications of decisions, and you'll always have a job. #### Related Items [[Artificial Intelligence]] [[Analytics]] [[Data]] [[Work]] [[Data Visualizations]]