You’ve bought an incredible concept for a weblog or LinkedIn article, if solely your inside material knowledgeable can discover time for an interview.
Human expertise and perception are key elements in high quality content material within the age of AI. Bots can churn out a great deal of content material at a speedy tempo, however with out human perspective, it feels generic.
The bottleneck is discovering the time to interview your CEO, CMO, Chief Content Officer or different inside SME who can present the point of view you want to elevate your content material.
We face this problem at SmarterX, which is what led the workforce to begin constructing an interview agent. It’s an experiment in progress however value sharing for you to take into account for your content material and advertising workforce.
Original Insight Is the Scarce Resource
As AI-generated content material turns into simpler and cheaper to produce at scale, the issues AI cannot replicate grow to be extra worthwhile: authentic quotes, firsthand expertise, robust opinions, perception earned by means of precise work and experiments.
At SmarterX, the content material workforce values an “expert-first” method to content material, the place the unique seed materials comes from a human, and AI assists in drafting and distributing.
That means repurposing human content material being created all through our group:
- Quotes and materials from our weekly podcast, The Artificial Intelligence Show
- Information from AI Academy programs
- Findings from authentic analysis
- Q&As with individuals throughout the workforce to floor information and perspective that AI cannot replicate
For that final one, Q&As, the issue is time and entry. Subject matter consultants aren’t all the time accessible when the content material workforce wants them or can’t allocate the time for an interview, even a brief, casual one.
The Experiment: An AI Interviewer
To resolve this, the SmarterX workforce has constructed a MVP (minimal viable product) of an AI agent that may interview consultants inside ChatGPT.
It works like this:
Let’s say the content material workforce has 5 posts deliberate for the week. For three of them, they need to incorporate authentic views from an inside knowledgeable. Rather than scheduling a gathering or sending an e-mail, the agent takes over the coordination and interview course of.
It follows these steps:
Research first. Before the interview begins, the agent researches the project, the subject, the viewers, and related issues the knowledgeable has mentioned in latest podcast episodes. It runs a number of analysis duties in parallel, so it is ready with actual context.
Adaptive interviewing. The agent then interviews the knowledgeable one query at a time. Each reply helps form follow-up questions. The aim is to draw out helpful, particular views not simply generic or surface-level takes.
Creates a usable temporary on the finish. When the interview is finished, the agent produces a structured temporary for the content material workforce: fundamental takeaways, pull quotes, a full transcript, and related supply context. It additionally fact-checks claims the knowledgeable made through the interview. Fact-checking issues as a result of even material consultants make errors, and constructing verification into the method saves the content material workforce a step.
But somebody nonetheless has to manually kick off the method, telling the agent what the posts are about and initiating the interview. The workforce is working towards automating this. For instance, having somebody tag a submit within the undertaking administration system as needing an knowledgeable interview, which might then immediate the agent to attain out and schedule one autonomously.
What It’s Actually Solving
The agent just isn’t a content material technology device however as a substitute is making it simpler for a busy knowledgeable to extra simply contribute their considering to content material. For instance, an SME may conduct their AI agent interview throughout a stroll or a morning commute, as a substitute of taking day out of their work day.
Whether this scales meaningfully, and whether or not the briefs it produces translate into noticeably higher content material, is one thing the workforce remains to be measuring. It’s an early take a look at of a narrowly scoped agentic workflow, one which appears achievable with no lengthy construct cycle.
The takeaway is that this: As AI adjustments the pace and scale of manufacturing, content material groups ought to be redirecting their power to making certain what’s produced is insightful, authentic and based mostly on what solely people can present.
This submit attracts on the AI Use Case Spotlight section of Episode 233 of The Artificial Intelligence Show, hosted by Paul Roetzer and Mike Kaput. To pay attention to the total Episode 228 of The Artificial Intelligence podcast, go to: https://podcast.smarterx.ai/shownotes/233
