When blogging first came to the forefront back in the early 2000s, it provided a way to deliver a complex idea to a broad audience, but more than two decades later could there be a way to reinvent that idea?
Kobie Fuller, whose day job is general partner at Upfront Ventures, thinks there may be a way to bring the blog into the modern age by taking advantage of the interactivity of generative AI.
Fuller’s idea centers around taking the notion of a blog post and making it a confined micro topic that users could explore by interacting with a bot. “Through the power of AI, what we are now able to do is take a standard long-form post and turn it into many different formats once we make the effort of constructing a sophisticated micro AI conversation that can adapt [to user queries],” Fuller told TechCrunch.
This can take the form of text-based conversational interfaces, or even AI simulated podcasts with AI interviewers asking the AI expert questions. Fuller has created a site called Kobie.ai to show what’s possible with this idea, while acknowledging it is very much in the experimental stage at this point, and he’s still learning what’s possible with this new format.
For starters, Fuller took a blog post he wrote for TechCrunch in 2014 on how emotion drives marketing, and ran it through GPT-4 with a very specific prompt, while using a chat bot to build the interactivity piece. “I layered a sophisticated prompt on top of the base piece of content as the sole source of information being utilized for the AI micro conversation,” he said.
That initial prompt has a couple of purposes. First of all, it creates an AI persona of the blog post writer, while providing structure for the conversation (as the conversation is led by the AI), making sure it covers certain topics and stays away from others. Fuller has created strict guard rails to keep the conversation on topic, and even includes a multiple choice quiz after the completion of the discussion to make sure the user grasped the topic.
He’s also been working on a podcast format, which was also created with GPT-4 along with his voice and the voices of a couple of actual podcast hosts. The result is an automated podcast. Although the voices sound robotic for now, it’s another way of interacting with the same source content in a different way.
Fuller acknowledges that this is a personal experiment, more than a business at this point, and that he would have to make it much easier to create original content and then transform it into interactive sessions or podcasts at scale, but he’s playing around with AI for now to help him understand it better, as he looks at startups in the space.
While the idea of rethinking the blog post is a good one, it doesn’t eliminate the need for the original long-form post. It just presents it in new ways that take advantage of generative AI capabilities. It will be interesting to see if this and other content format experiments we’re seeing will eventually turn into businesses that would be backed by people like Fuller and his firm.