Episode 262
Avoiding AI Slop: Best Practices from the Human Framework
AI is transforming how we create content, but there’s also a growing problem: AI slop. It’s the flood of auto-generated videos, images, and text that takes up space without adding value; and we’re seeing it everywhere. So how do you use AI to enhance your work without the noise?
Host Matt Pierce introduces TechSmith’s HUMAN Framework, a practical approach to keeping the human element at the center of AI-assisted content creation.
We hear why your expertise and lived experience gives you access to nuance that AI can’t replicate, how authenticity has become currency in content creation, and why treating AI as a thought partner rather than a replacement leads to better outcomes.
Matt also shares TechSmith’s philosophy on AI integration, including their approach to control and customization in tools like Camtasia AI. While AI is the worst it's ever going to be (meaning it will only get better), the choices we make now about how we use it will determine whether we're creating valuable content or just more noise.
Learning points from the episode include:
- 00:00 - 02:01 What AI slop is and why being able to identify it is crucial
- 02:01 - 02:28 TechSmith Unlocked event announcement
- 02:28 - 03:23 Introducing the HUMAN Framework acronym
- 03:23 - 04:40 H - Harness your expertise
- 04:40 - 05:38 How humans have access to nuance through lived experiences
- 05:38 - 07:47 U - Understand your audience.
- 07:47 - 09:00 M - Make it authentic, not artificial
- 09:00 - 10:31 A - Aim for better, not just faster
- 10:31 - 11:52 N - Never skip reviews
- 11:52 - 12:30 How to use the framework as your guide
- 12:30 - 13:00 Why you need to choose tools that keep you in control of the output
- 13:00 - 13:28 The importance of ethical AI partnerships
- 13:28 - 15:40 Control and customization in Camtasia AI
- 15:40 - 16:30 AI as an augmentation tool
- 16:30 - 17:17 Outro
Important links and mentions:
- TechSmith Unlocked: https://discover.techsmith.com/techsmith-unlocked-2025/
- Camtasia: https://www.techsmith.com/camtasia/
Transcript
Good morning, good evening, good afternoon, wherever you are and wherever you're watching from. My
Speaker:name is Matt Peirce, host of Visual Lounge, and today I'm going to be talking
Speaker:about something we like to call a I Slop. That's right.
Speaker:It's a phrase that's getting thrown around a lot and it's got some meaning. I
Speaker:want to start off with a little bit of a story here. It's just going
Speaker:to be me today, but so we'll have some more great guests coming soon. So
Speaker:here's the story I want to tell. So I've been working with a variety of
Speaker:different groups, including our PR group, and we wrote an article together about
Speaker:AI slop and some things that are happening on channels like YouTube
Speaker:and other places that are kind of trying to combat AI
Speaker:slot. Because there's a lot of stuff going out, videos and music and
Speaker:just audio that's, you know, it's just auto generated and it
Speaker:has really not as much meaning or purpose or
Speaker:it's really not doing much, just taking up space and not adding value to their
Speaker:systems. Obviously a concern. So we wrote this article talking
Speaker:about kind of the changes, the impact, things like that. And it was
Speaker:funny, within one hour of that article getting published on a reputable
Speaker:site, someone had duplicated it using AI
Speaker:making changes, and it wasn't nearly as good. Now,
Speaker:crazy, right? Like the things that AI is enabling people to do, to move
Speaker:quickly. Obviously it can be a generator,
Speaker:something that can speed you up. It can be something that can add value, to
Speaker:do things that maybe you couldn't do otherwise. I think about like
Speaker:vibe coding, but if you're in the world of marketing,
Speaker:learning, design, training, you're maybe even in
Speaker:product management or whatever it might be. There's a
Speaker:lot of things getting thrown at us right now about AI,
Speaker:and it can often be challenging to know, is this a good use?
Speaker:Is this a bad use? Is it, you know, is this really good
Speaker:content? Is it bad content? And so today, if you're
Speaker:listening to this, on the day the podcast October 1st is being released, we are
Speaker:hosting an event, TechSmith Unlocked, where we're going to be talking
Speaker:about our vision for AI. And part of that event, one of the
Speaker:things we were going to do is roll out a framework. Now,
Speaker:time constraints, got a lot to show you, a lot of cool features, not
Speaker:all of them, all AI things, but a lot of cool things coming down the
Speaker:pipe that you can go sign up for unlocked if you haven't, or you
Speaker:can always listen, get the replay, we'll be sharing out bits and pieces as well
Speaker:as the full two and a half hour webinar. Let me share this about
Speaker:what we were going to share. We're going to share what we're calling the Human
Speaker:Framework. And the Human Framework is an
Speaker:attempt to start to look at how you
Speaker:apply AI, whether you're making video recording
Speaker:audio, you're doing images, whatever it is, you could apply this to
Speaker:that and to give you some guidance and direction in the time
Speaker:of kind of this AI Wild, Wild
Speaker:West. And now I want to be clear. I realize fully
Speaker:that AI is, as they say, the worst it's ever going to be.
Speaker:Its usefulness now may be limited compared to that
Speaker:which it may become. AI
Speaker:may not have yet realized its full potential. Right.
Speaker:It has not become actualized as far as I know. But the
Speaker:thing that I think is important is that as creators, as
Speaker:people are creating content, we're thinking through what that might
Speaker:mean. Now I want to go through what human stands for, because
Speaker:it's an acronym, right? H U M a N. And then we'll
Speaker:talk about each of these really briefly. So first is you need to
Speaker:h harness your experience. Then
Speaker:u understand your audience, because that's
Speaker:something you can do. M then make it authentic,
Speaker:not artificial. A aim for
Speaker:better, not just faster. And n
Speaker:never skip reviews. Now we're going to go through each of those, but I think
Speaker:it's important to think through, like, you know, these pieces and we picked them
Speaker:for a reason. There's probably other advice, but hopefully this becomes something that you can
Speaker:turn to. Say, like, hey, I'm using AI. Am I
Speaker:keeping it human? Have I harnessed my expertise? Understand the
Speaker:audience, understand am I making it authentic? Am I aiming for better, not
Speaker:just faster? And am I keeping, you know, not skipping my
Speaker:reviews? So first, with harness your expertise. AI is meant
Speaker:to be general. In our past podcast with Andy Cadena, we talked very
Speaker:specifically about AI is going to aim for the common
Speaker:denominator, like the kind of the middle ground and everything. And that can be really
Speaker:good, especially if you're like, I don't know much about that, or I'm not really
Speaker:an artist. I don't know how to draw an icon. Maybe it can make an
Speaker:icon for me. So it's aiming for the middle. But you have expertise
Speaker:as a human. You have experience, you have access to nuance,
Speaker:nuance that is capable because you have lived through
Speaker:experiences that you might not be able to read. I won't go deep into this,
Speaker:but there's a lot of pieces about various Things that are happening with AI because
Speaker:people are turning kind of it over AI over to be their therapist
Speaker:or to give them advice, and they're, they're choosing to make decisions. And
Speaker:one great report I read talked about that nuance is really, really, really
Speaker:hard, even for humans sometimes. And AI just at
Speaker:this point does not get it. So it doesn't understand that you're
Speaker:hinting at something or thinking about something or, or there's gray areas.
Speaker:It's like this is the answer and that's it. And that can
Speaker:be problematic. And when it comes to like learning development, it comes these
Speaker:other opportunities. We want to make sure that your judgment, your expertise, your
Speaker:understanding is getting applied. Next is understanding
Speaker:your audience. Now, AI also referring back to the episode with Andy
Speaker:Crestedina, AI is really great at helping you
Speaker:to maybe think about your audience in a different way or to start, provide you
Speaker:information about your audience. But only you can choose the audience
Speaker:and only you can make sure that that audience actually
Speaker:matches up with your intention. So you might ask
Speaker:AI, for instance, to look at your website, give you
Speaker:kind of a portfolio view of who it thinks your core audience is.
Speaker:Great. But is that actually who you're talking to, that who you're making
Speaker:your training for? Is it really their lived experiences that. So you
Speaker:need to pull those in. And again, I can augment here, can enhance it can
Speaker:help, can you can use it as a thought partner to ask it questions and
Speaker:have it challenge you. But at the end of the day, you have to make
Speaker:a decision about your audience, what they need, who they are
Speaker:and how you can help them and in the best possible ways or
Speaker:better to connect with them. So, you know, asking yourself, who's this really
Speaker:for? What do they need to know? Are my really giving
Speaker:that to them and then really digging in a little bit deeper than
Speaker:what? Again, nuance that the AI isn't going to be able to get you to.
Speaker:And again, it can help cycle up that a little bit faster, a little bit
Speaker:more, maybe think more deeply. But you've got to make some judgments
Speaker:there. The next in M make it authentic, not
Speaker:artificial. Authenticity
Speaker:cannot be faked. AI can't do it. It's getting there. It's going
Speaker:to be harder and harder as we go. But I think being
Speaker:authentic, speaking from the heart, is something that it's going to have a hard time
Speaker:doing for a long time now it can start to replicate ums and ahs,
Speaker:pauses, mistakes even, maybe.
Speaker:But is it going to be able to modulate its voice appropriately? Is
Speaker:it going to Be able to do things that a human can do. And in
Speaker:our imperfectness, right, in our frailty, in our ways that
Speaker:we are sometimes terrible or bad or not good,
Speaker:that becomes a calling card. Not to say we should make bad stuff, but just
Speaker:the reality that like I lean in or I back
Speaker:away or I do things that maybe aren't the
Speaker:way perfectly done, but that has value in your
Speaker:content. Now, I'm not saying going out and make mistakes or, you know, do that
Speaker:just on purpose, but be re. That authenticity, it's. It's a real
Speaker:currency. We can trade that in for the content, you know, help the
Speaker:people that are, we're trying to connect with, right? So that comes down to the
Speaker:choices that we're making, you know, and sometimes it's going to make sense to go
Speaker:with an AI avatar or an AI voice or a stock
Speaker:visual or something, right? But like, let's make sure
Speaker:we're making those decisions with purpose, right? So that we're
Speaker:not just letting it be like everything. Because some things
Speaker:really matter. Like think about the experiences that you want to have
Speaker:and where. Where it makes sense, right? You know, if an AI
Speaker:avatar was the one to explain that your benefits are
Speaker:getting cut or you're getting fired, or maybe
Speaker:it's explaining the value of something else that you're like,
Speaker:that's meant to be personal. It's meant to be human to human. And
Speaker:those are. Because those are hard things and emotional and AI does not care.
Speaker:It will say whatever you ask it to say. Essentially. Now,
Speaker:within limits, of course, next, aim for better, not just faster.
Speaker:Of course, AI can speed up a lot of the process, but this is
Speaker:where again, you come in as a human. When
Speaker:we think about video. I've been playing with VO3. We talked to Ellis Pratt
Speaker:recently about VO3. It can make a lot of cool stuff. You can do
Speaker:a lot of things with it. But is it better?
Speaker:It could be faster, but is it better? And you know, right.
Speaker:What I find right now is I'm dealing with a lot of, like, the
Speaker:quirkiness that it outputs. And it's going to get better. It's going to get better
Speaker:and it's going to get easier to understand. And I could get better at writing
Speaker:my prompts and doing the things that I knew need to do, providing
Speaker:context. But, like, what we're really trying to do is enhance the things that we're
Speaker:doing now. Even though I said hey earlier, AI
Speaker:aims for the middle, which it does, it might be able to
Speaker:help elevate you because I find Often again, I like
Speaker:the, again this example I've shared, but like, I use it as a thought partner.
Speaker:I'm saying like, hey, I'm thinking about these things, you know, one by one,
Speaker:like one at a time. Ask me a follow up question to help me to
Speaker:get to, you know, to whatever. And that clarity is like having
Speaker:a conversation with someone really.
Speaker:AI maybe writes better stuff, but I write better stuff
Speaker:because of it. And I'm more thoughtful about the things that I'm putting down on
Speaker:paper or the video that I'm creating or whatever it is that I'm doing
Speaker:because it's helping me to get to a better conclusion. And
Speaker:that's not always faster. It could write something
Speaker:one take. It could absolutely write something, one take and be done.
Speaker:And it would be done, but it wouldn't be better. Okay, the last
Speaker:thing is the end. Never skip reviews. We
Speaker:know AI makes stuff up. It
Speaker:hallucinates this term. It's gonna do it. So it's
Speaker:critical that we're reviewing whatever it's putting out. There is a temptation because it
Speaker:did something we just gotta get done. Maybe it's not that important that we don't
Speaker:review it. You've got to review it. We've gotta save
Speaker:yourself some embarrassment. Save yourself
Speaker:and your learners or your audience confusion.
Speaker:The worst case is that you're doing it for something that's really critical
Speaker:and it makes a mistake and someone gets hurt, that's not good.
Speaker:The best case is that like, oh gosh, that's dumb that it said that.
Speaker:Ugh, let's just, you know, whatever. But even in that case,
Speaker:what's the trust that's happening, right? When are you willing to erode that trust?
Speaker:And so we just need to be thoughtful. Look, I get it.
Speaker:Sometimes it's just fast, let's see what it cranks out. It's going to help you
Speaker:move along, but take the time, right? Like take the time
Speaker:to just go through it. Especially if you're going to share it with
Speaker:other people. You know, even with a, like, say
Speaker:there's different AI tools out there. If it's a caption that maybe is a little
Speaker:bit wrong, I'm sure, maybe even on this podcast, just being thoughtful about
Speaker:the impact that has on someone who might need that content,
Speaker:who might be relying on it and you might be thinking, oh, it's just, I'm,
Speaker:I'm glad I'm doing it as bonus. But
Speaker:we want to just be really careful about that, right? So
Speaker:all in all, with a human framework, it's a chance to stop and
Speaker:reflect about the things that you're doing, the ways that you're using AI.
Speaker:There's AI is out, the Pandora's box is open. It's
Speaker:not going away. Tools are. We're going to see a continued rapid
Speaker:rise of tools. I believe that we're going to see more and more tool sets,
Speaker:more and more options of things that can do stuff and that they're going to
Speaker:get increasingly more complex, increasingly more
Speaker:sophisticated. And so we have a right to, you know, we need to do some
Speaker:due diligence. So, you know, as you're going out seeking
Speaker:tools that can do AI things,
Speaker:make sure you're choosing the things that are going to allow you to remain and
Speaker:keep the human in it. Both kind of from this
Speaker:framework, but also just kind of naturally as well. Right? Like
Speaker:be in the driver's seat still, maybe there's a day
Speaker:that AI is so good that we don't have to worry about that and we're
Speaker:all living in a Shangri La where we don't have to work so hard.
Speaker:But until that day, it's critical that you are part of
Speaker:the work that is going forward. You're and using AI.
Speaker:And that's not saying don't use AI. I think you got
Speaker:to embrace it. But make sure you're choosing tools that
Speaker:are ethically sourcing their materials, that are partnering
Speaker:and keeping you secure and thinking about that long term future.
Speaker:Of course we want them to make better models and train things and we want
Speaker:them to create things that make our lives easier. Who doesn't want
Speaker:that? But make sure you're, you're thinking about that,
Speaker:right? We believe at TechSmith that there is
Speaker:a level of control and
Speaker:customization that should be yours. So for instance, if you haven't heard
Speaker:Camtasia AI, it's a Labs project for us.
Speaker:We've taught basically an AI to create screencasts from images.
Speaker:So it's pretty darn cool. But one of the things that
Speaker:our designers have built into it is a lot of control. So even be,
Speaker:even if you don't go into Camtasia to change everything, you can change the focus,
Speaker:you can change the script, you can do lots of adjustments
Speaker:before you ever produce a video. An MP4, you could produce just an MP4,
Speaker:but you still have a lot of control. Or you could take it into Camtasia
Speaker:and you can control all of it. It's all built
Speaker:into a Camtasia project and that kind of control
Speaker:puts you in the driver's seat. Now, there might be videos you're like, I don't
Speaker:want to change anything. It's fine, it's good. It's good enough. The context,
Speaker:right? Maybe I got to change a word or two, whatever, but it's good. And
Speaker:there's my times. You're like, this got me a first good start. But I'm going
Speaker:to go change the heck out of it because it's not where we want it
Speaker:to be. And we think that there's a lot of power in control and
Speaker:being able to, to adjust, right? Like when you put in what you put
Speaker:in a prompt, if you just get back the final solution answer, it's like,
Speaker:you know, now I gotta, you know, I, I've had very few
Speaker:videos come back from other tools that have allowed me
Speaker:to just use it without then having to take in, edit it or change it
Speaker:or find some other pieces to put together over it. So
Speaker:ultimately, at the end of the day, you've got to find a balance there. There's
Speaker:not a one solution here. There's not one
Speaker:AI. You know, we believe that partnering
Speaker:with some of the best in class, best of breed, however you want to phrase
Speaker:it, is going to be really powerful. We have announced in other
Speaker:places that some of our tools, we're integrating with 11 labs,
Speaker:we've got other tools that we're using and pulling in that we feel are
Speaker:some of the best that are going to power some of the AI
Speaker:technology behind what our tools do and then yet
Speaker:allowing you still to have control. So we feel very strongly
Speaker:that you need to be in the driver's seat. You are the human, you are
Speaker:in control. And this framework hopefully will help you wrap your mind around
Speaker:some best practices, things to help you. Look, we're not telling you how to prompt
Speaker:things. There's lots of people out there tell you how to prompt things, not telling
Speaker:you all the details. We have other stuff that we, other conversations we had about
Speaker:building things. But this is about centering yourself
Speaker:as a creator around how are you going to interact and use
Speaker:AI as an augmentation, as a, as a building
Speaker:block to do the things that you need to do. And, you know,
Speaker:maybe this is a hope, this is an okay metaphor, but like, think about
Speaker:when you augment your, you put on glasses, you see a little bit clearer, or
Speaker:if you needed a prosthetic for something that helps you and enables you, or if
Speaker:you use things in your shoes, any kind of augmentation, right? It enables you
Speaker:to do more, maybe better, faster, more capably. And that's
Speaker:what AI is doing, right? Now and we can use it well, but also
Speaker:it can hinder you. So just be mindful about that process. But again,
Speaker:let's wrap up with human Harness your expertise,
Speaker:understand your audience, make it authentic,
Speaker:not artificial. Aim for better, not just faster.
Speaker:And never skip reviews. With that said, I hope you
Speaker:found this unaugmented version of the Visual Lounge helpful and that
Speaker:I hope that the human framework is something that you can apply. You can watch
Speaker:for coming soon from us, there'll be a PDF with an article
Speaker:blog post. We've got some other things that we'll put together to help emphasize these
Speaker:ideas, but I wanted to share it here first in light of our TechSmith
Speaker:unlock event. And if you haven't again TechSmith unlocked, you can
Speaker:go search for that. We'll have stuff all other places you can find.
Speaker:If you're missing it, I apologize about that. We'd to love love to have you
Speaker:join us and if you can't again can't watch it. Don't worry,
Speaker:we will be sharing out that content. We got a lot of cool things coming,
Speaker:some fun announcements, some other things that are changing up. Watch out
Speaker:moving things forward in a good, good way.
Speaker:So with that said, we love our audience at the Visual Lounge. So
Speaker:grateful for each and every one of you taking time to listen. Give us a
Speaker:little bit of your attention and we'll what we hope is that you learn and
Speaker:grow and continue to be better by spending just a little bit of
Speaker:time leveling up every single day. Thanks everybody.
