Commentary

Do Humans Interpret Context Better Than Robots?

I'm beginning to wonder whether robots can interpret content better than humans, as Google, Bing and others build platforms that teach humans how to team them. Today, however, we're not here to talk about search. We're at OMMA LA to talk about robots and programmatic creative.

Jim Sterne, cofounder of Digital Analytics Association, led the discussion with panelists Zach Glass, VP of digital advertising and social at Red Interactive Agency; Matthew Nally, founder and managing partner at Labmatik; Zachary Soreff, president at Sawyer Studios; and Steven Wong, co-founder and CMO at Ready State.

The panel debated whether teams can hand the creative process off to computers and whether this would eliminate the creative department.

Wong said many of the aspects required to create creative pieces can be "handed over" to algorithms, but Glass believes some of the processes need a "human touch." In fact, he said, humans are needed to test the context to ensure that the measurement teaches marketers something about the process.

Soreff said automation will do all the boring stuff. He suggests testing different lengths and edits of videos, mapping about 200 variables. Humans are needed to test the clusters of data that marketers deem less important to certain audiences, although he trusts the machine to kick out the best advice to achieve the best result. The machine will likely suggest an audience that marketers had not planned to target.

I've actually heard that many times from a variety of marketers who say programmatic platforms typically map the campaign or context of the content to a specific audience they had not thought to target. 

The machines can crunch the hard numbers, but humans focus on the intellectual part of the message and campaign.

How far into the future will we go before machines can figure out the human part of the equation?  I believe it's beginning to happen now. 

Next story loading loading..