The case in point is the availability of information the power to process that information and how much of both is needed to make profitable decisions. Sane decisions can be made with relatively less information but that is at the cost of risk. In absence of accuracy the risk must increase and the processing unit, that is your employee, must rely on his/her experience. The purpose of data and algorithms is not fancy charts and stunning visualizations, it is to provide simple insights that reduce risk by augmenting experience, and sometimes even correcting it.
It depends on use case but most non life threatening business situations require only a certain accuracy to be meaningful. The data doesn't improve the verity, but the processing and tuning the machine, whether it be the functional areas of your organization, warehouse operations, website optimization, speech in sales calls, transcripts in chats, features in your products or thousand nuances of your various customer touch points, or the effect of a light spectrum on every leaf of a plant in a hydroponic farm, does.
The democratization of static and iterative insights and automation of required actions, build upon the scalable computing stacks in cloud and out of the box algorithms that fit various business cases, will define future competition in a different way. Yes the employees will remain an asset, ignore the attention grabber title, but as the availability of data and ability of machines to crunch and understand the data increases so will the competition and it will test the employees ability to harness data and ingest the real time recommendations. An understanding of how the organization can augment the information processing power of its employees will create scalable advantages for the long term.