As AI is deployed more widely, business leaders will learn the hard way that humans need to be intimately involved in every use case to ensure effective safeguards and oversight. This will be familiar to companies that went through cloud migration, as opposed to those native to it. Just a few years ago, many companies that migrated to cloud-based infrastructure with little to no strategy saw server sprawl become virtual machine sprawl, which then became cloud sprawl. They lost control over essential components of their business and learned the necessity for developing rigorous oversight.

Another past lesson that can be applied to AI concerns data lakes, or centralized repositories for structured and unstructured data. As data grows ever more important, organizations will need to refine their approach to data integrity, purpose, and access. I predict that we will see a flood of new tech companies that specialize in all the unglamorous uses of data—such as labeling, validating, and structuring—that make AI possible.

For all this, I’m very optimistic. I believe we’ll look back on 2024 as the year we deployed AI not just to help do our current work better, but to pioneer entirely new and transformative ways of working. That’s a fundamental change that can’t come soon enough.