In the last year, everyone has been talking about how AI is going to change the way people work and replace jobs normally performed by people. But like so many technological innovations before it, AI will do more to create opportunities for the people that know how to use it.
At Penta, we invest in understanding and testing what AI can do well, and where it falls short. We know the best possible system is one that combines human expertise and flexibility with the speed and scale of AI.
Artificial intelligence, and specifically the new breed of generative AI based on large language models (LLMs), can be powerful when used in combination with experts who know both how to harness them and overcome their limitations.
Generative AI is based on language models as inputs, and while massive inputs may seem to make it more accurate, it can also lead to problems if you are basing the output on data you don’t see and can’t access.
The way a text-based generative AI product works right now is that you provide a prompt like “Who is the first person to step foot on the moon?” and the AI looks at the entire corpus of words and content in its training data, and determines what is the most likely next word in that sequence. Then it does that again and again. The words have the appearance of coherence because it is based on existing language, but only an incidental sense of truth. A generative AI does not know who first stepped foot on the moon, only that the highest probability next word in that sequence is “Neil.”
This creates a significant hazard– the AI does not know what is true, and basing responses on probability leads to hallucinations (asserting facts that are objectively false) and relying on the most common answers.
Many media monitoring companies are using the AI buzzwords to promise accuracy that they cannot guarantee, using tools that are not transparent to put more barriers between the technology and the value.
We know firsthand the value that human curation brings to the media monitoring process. As communicators, we have used every system and process to make sure we know what is happening as quickly as possible.
We need analysts who are responsible and accountable for making sure the alerts are relevant, using all the technology available to deliver the highest quality alerts every day.
We believe that human curation is essential for effective media monitoring. AI can provide a broad overview of the media landscape, but it’s our human expertise that allows us to provide our clients with the insights they need to make informed decisions.
If you’re looking for a media monitoring solution that provides the accuracy, flexibility, and responsiveness you need, then you should consider a solution that incorporates human curation. With human curation, you can be confident that you’re getting the most out of your media monitoring investment.