When I first began working in digital, we called it the Wild West. No rules or guidance were directing the field. As the complexity of digital programs grew, so did the sophistication with which we managed them. Now, with the meteoric rise of AI products and services and the commonality of ChatGPT, we face a similar frontier.
At the Forefront of AI
As an AAPC board member, I’ve been at the table in developing an industry-wide code of ethics around AI and promoting a standard of excellence for using new technologies in this area. In a recent session at Campaign Tech, my colleagues and I discussed the pluses and minuses of AI in digital politics. Here are my hot takes:
Benefits
– AI can provide excellent targeting models. A tool whose whole value proposition is learning through data would help political operatives create the next level of sophisticated and unique data sets, models, and more for digital advertising targeting.
– AI can improve customer service by serving instantaneous responses when customers/voters are looking for information such as: Where is my voting location? How can I get a ride to vote? AI chatbots can help candidate websites guide voters to precisely the information they seek and cast a wider net for education.
Areas of Concern
– We are concerned, however, that relying on AI will preclude building a significant bench of expert Digital practitioners. Will our teams be able to learn instinct and judgment when they are continually relying on AI?
– How are we addressing the security and privacy concerns of putting this information into a database?
– AI language models are known to “hallucinate” and provide flat-out wrong information from time to time. Without sophisticated guidance, lazy digital practitioners will miss these mistakes and contribute to a terrible political culture of disinformation.
– Will we all start to sound the same? With crowd-sourcing and GPS information, when everyone listens to Waze after a crash or traffic incident, don’t we inevitably cause another traffic backup on the new route if we all use it until someone else veers in a different direction?
– AI language learns from the data they are trained on. Since there is so much digital content that is low quality, AI will reproduce the same flawed strategies endlessly without sophisticated strategists there to guide it.
AI in Practice
Though AI is buzzy as of late, we have been implementing analytical AI expertise into our award-winning targeting programs for a long time. As the technology evolves, so do sour tactics so that we deliver the best in the industry targeting performance for our clients.