Now that we’ve pretty much established that chatbots are really, really stupid, it looks like the next big thing in AI will be something called the intelligent agent. What that means, in theory, is that AI will take on different roles for different purposes and, in the end, create content and work that is, also in theory, better than what you’ll get from a regular chatbot.
“In theory” is obviously doing a lot of heavy lifting here.
For one time only, our upcoming AI Fundamentals class is going to be focused exclusively on PR work. We’ve 100% revised the class to take into account all the advances in prompting, apps, and techniques from the last few months. We’ve also curated a custom toolset of apps specifically crafted to speed up PR writing, pitching, and creating on-brand media (think: articles, images and videos).
The class is happening June 20, and spots are limited. If you reserve yours now and use the code AISPRING at checkout you can get a 50% discount. If you’re new to AI, or if you’ve plateaued in your use of it in your work, this class will take you to the next level, and give you knowledge that you can bring to your colleagues. Like I said, this is happening one time only, so don’t wait!
The industry has been talking about intelligent agents for a while. Google released an article in 2022 describing what amounts to the definitive implementation of the concept. Basically, it says that different AIs will do different parts of a task and the results will be better than if a single chatbot does the task.
Think of these agents as parts of a content assembly line. As the content moves through each agent, it grows and changes. The agents create, refine, and ship the content and you optimize each agent to the task at hand.
For example, in the context of writing a news article, one agent will be optimized to write an outline, another will be optimized to write the story, another will run a fact check, another will copy edit. An “editor-in-chief” agent will then come along and make sure the voice and quality are intact.
The same thing is happening in marketing. For cold calls, for example, one AI will figure out a potential lead, the next one will analyze how to attract that lead, and a final AI will make the call and try to convince the person on the other end to listen to a bot.
If you’re paying attention, you’ll notice that in both of these scenarios the human is left out of the equation. By splitting the AI processing into agents you reduce the chance of hallucinations. But AI is AI and if the outlining agent hallucinates than the editor-in-chief agent had better know that glue doesn’t go into pizza or the whole thing is ruined.
Agents have been useful in translation primarily because the steps required between translating text and making it human readable are fairly simple. But agents are now editing code and making bill collection calls and eventually they’ll be doing even more interesting and important things. Just be ready to tell your copy writing agent to shape up if it starts telling us to eat more rocks.
Leave a Reply