With AI now able to edit content really well, how would you create an offer for content repurposing, how would you charge a business, and would you target businesses with no marketing exposure or ones with podcasts you could repurpose at volume with AI? How would you differentiate from existing agencies that aren't using AI?
Simple and straightforward: I would have AI break down long-form clips into short-form clips. I would target preferentially long-form platforms like LinkedIn, full-length X videos and X podcasts, and obviously YouTube videos like mine. I would build three major systems: a video-to-text-post converter, a video-to-static-image converter, and a video clipper. Video-to-text-post converter is pretty straightforward — we've been able to do that for a very long time — you take content that resounds within the transcript of the main piece of long-form content and cut it up for X, Instagram, whatever else. Video-to-static-image is basically the same thing, but there are really strong formats right now where people take what is said in a video transcript, cut it up into really high-ROI segments — the best parts — and turn those into notes that look handwritten. Ormozy did this, a bunch of other guys do this, it's a very common thing to do. Video clipping is where you take the transcript but also identify silences, and use both of those to feed into an AI model to cut at specific moments. You can then add effects, make the card half as tall, put yourself in the bottom half and have visuals and research on the top half. AI can actually do all this stuff now basically completely autonomously.
Yes, of course. I recommend refactoring your internal flow that generates these clips. Video processing doesn't need to take 70–90 minutes if you parallelize. For heavy workloads, people often rent a high‑core VPS (not a full supercomputer) with fast network storage, using saved configurations or code that pulls the video quickly. This lets them parallelize, run the jobs, and finish in about 5 minutes instead of 2 hours on a personal computer. To scale, turn this into a real podcast offer that saves time per client—it's easy and simple. Besides that, focus most of your time and energy on outreach, as most people do. Nick notes that there's still an overhang between what people think AI can do and what it actually can't.
My main workflow is repurposing my daily updates onto four or five social media platforms. I’m only using a small fraction of the possible channels—there are about ten other places I could publish. If I built a full publishing pipeline, I’d double my current results. Currently, I record two video streams: one with my face and a comment card (useful for some channels), and another with just my face. I use AI (Whisper) to generate a transcript, then clip the useful chunks, create Q&A content, and also produce straight‑answer videos. I publish these across Instagram profiles, YouTube Shorts accounts, and similar platforms. Feeding the transcript into a model like Astra would perform reasonably well.
No, it is not. Thank you. We did try AI‑generated content. I should note the stack so far seems to be: use HeyGen to replicate the person (that’s the video), then use 11 Labs to replicate their voice (the audio). If you really want to add "alpha," you’d add a noise layer and alter the noise levels. The reason is that a lot of AI‑generated voices are just the voice with no other audio, which is very noticeable. Adding background noise and other ambient sounds makes it less noticeable. We tried that on Instagram, but honestly we didn’t do a very good job and I’m not sure the video is still up, so I’ll probably remove it. We keep experimenting, but I feel a much better approach is to be transparent—clearly label the content as AI, e.g., "XYZ AI," or use very short‑form platforms where repeatability isn’t an issue. On YouTube long‑form, everyone knows who you are, so AI‑generated content stands out more. Short‑form TikToks for a faceless channel can work as AI‑influencer content.\n\n"Okay, anyway. Eric says, \"This pushes me to actually accomplish the things that I should be doing sooner in my business and will initially drive me revenue and growth.\" Thanks very much, Eric.",
You know, I actually did this exact same thing using a service called Be Human a while ago. I'm not affiliated with them and don't use their service anymore; I'm not even sure if they're still around, but it was a pretty cool service. You could record a video, say the person's name in it, and the AI would find that point, blur over your mouth, and use a training sample of your voice to say the name. For example, I'd say, 'Hey, really cool seeing you. Thought I'd record a quick video just putting this in front of you to see if you get some value out of it.' The video would then say, 'Hey, Peter.' The issue back then was that the models weren't good at distorting your face, so my hack was to grab the Loom video recording bubble (or equivalent) and say, 'Hey, Peter.' While the video was moving fast, you couldn't see my face, and if someone paused on the name to double-check, they'd see me halfway through a transition, making it unclear. Some of the best sales and lead gen cold email campaigns a couple years ago used this approach and flew under the radar. You could create and record the video, give it total plausible deniability, make it seem super customized, and then pass it through. If you're a hacker nerd wanting to build your own pipeline, you could do this for about 5 cents per 200 videos by renting GPUs on a service like RunPod, running a video name‑change model, or using whatever model Be Human uses, treating it like an API point. So yeah, I'm not going to discourage you.