Proprietary AI

AI will always have a level of general use and applicability, such as tools like ChatGPT and Bard, but what most organizations will end up with, as they should, is process-specific, proprietary AI.

So what does process-specific, proprietary AI mean? It means AI on top of models leveraging an organization’s data for operational areas of the company, such as sales, marketing, customer support, product management, and more. Organizations may augment their models with outside data to help train or evolve them faster for more efficient, accurate, and valuable output. But rest assured, the goal for most organizations will be to have a locust of control over their proprietary data while leveraging machine learning and artificial intelligence.

More and more open-source models are being created and made available to organizations to leverage. Open-source models will make sense for some organizations, while it will make more strategic and operational sense for others to stand up and manage their models or some hybrid. As the cost of compute processing comes down, the ability of organizations in the mid-market and nonprofit sectors to operate and maintain their models will be increasingly viable. In a subsequent post, I will write more about the sector-specific ML and AI impacts, but I want to stay focused on the growth of proprietary, task-focused models and AI.