How to measure what a human actually added to a piece of content — above what AI would have written on its own — in 11 minutes.
Watch the Episode
Having a team that understands and can implement AI-based tools and approaches is critical. But how do you know if your team is AI-ready? One answer starts with the work itself: if AI can produce a competent version of anything, the only thing worth measuring is what your people add on top of it.
Welcome to One Amazing Thing About Hupside. Today I’m talking with Jonathan Aberman, Founder and CEO at Hupside, about Hupmapper — a tool that scores how far a piece of writing goes beyond what AI would have generated on its own.
What you’ll see in this demo
Hupside is a company building tools that measure human originality in an AI-saturated market. In this episode, Jonathan Aberman, Founder and CEO at Hupside, demonstrates Hupmapper, a product that takes a piece of written content, compares it against a baseline of how AI would have generated the same content, and returns an originality intensity score along with a unique visual fingerprint for that sample. Unlike an AI detector, it does not try to determine who or what wrote the text — it measures how much value was added beyond the AI baseline. It’s most useful for marketers, consultants, editors, and educators who need a defensible signal that a piece of work is worth reading.
Key takeaways
- Most AI-detection tools try to establish whether a machine wrote a piece of content; Hupside’s argument is that the more useful question is where a human added value beyond what AI would have produced.
- Generative AI functions as a similarity engine, so originality above the AI baseline can only come from a human working alone or a human working with AI as an augment.
- Hupmapper analyzes a piece of written content against a baseline of how AI would have generated the same content and returns an originality intensity score.
- A score of zero does not mean AI wrote the content — it means the content is indistinguishable from what AI would have produced anyway.
- Hupmapper accepts up to 2,000 words of pasted text or an uploaded file, and returns both a numeric score and a unique visual fingerprint for that sample.
- Because Hupmapper compares content against an absolute AI baseline rather than looking for stylistic tells, it is not subject to the false positives and cultural bias that affect conventional AI detectors.
- Originality of authorship asks who made something; originality of output asks whether the result exceeds what AI would have produced. Only the second one carries economic value.
- Originality and quality are not the same thing — content can be entirely original and still not be good.
- Practical uses include a consultant demonstrating that deliverables exceed AI output, an editor triaging which submissions merit review, and a reader deciding whether a piece of published content is worth the time.
About Hupside
Hupside is a company measuring human originality in AI-integrated work. It created the Original Intelligence category and builds two products: Hupmapper, which scores how far a piece of written content goes beyond a generative AI baseline, and Hupchecker, an assessment that measures an individual’s originality and returns an Original Intelligence Quotient (OIQ). Hupside closed a $1.7M pre-seed round led by Ruxton Ventures in 2025. Learn more at hupside.com.
About Jonathan Aberman
Jonathan Aberman is Founder and CEO of Hupside, which he started with a group of scientific co-founders after roughly 25 years in venture capital — he left the fund he was at when he encountered the university research the company is built on. He is a partner at Ruxton Ventures and served as founding dean of the School of Business and Technology at Marymount University from 2019 to 2023. He has been named a “Tech Titan” by Washingtonian, listed in the Washington Business Journal’s Power 100, and recognized by Virginia as one of its 50 Most Influential Entrepreneurs. He is the author of the forthcoming book The Originality Dividend. Connect with Jonathan on LinkedIn.
Frequently asked questions
How do you measure originality in content?
Hupside’s approach compares a piece of writing against a baseline of how a generative AI model would produce the same content, then scores the distance between them. The result is called originality intensity: the higher the score, the more the piece contains that AI would not have generated on its own. A score of zero means the content matches what AI would have written anyway.
Is Hupmapper an AI detector?
No. AI detectors look for stylistic tells that suggest a machine wrote the text, which is why they produce false positives and show bias against non-native writers. Hupmapper does not attempt to identify authorship at all. It measures the gap between a document and the AI baseline, which is an absolute comparison rather than a probabilistic guess about origin.
How do you tell if content is AI-generated?
Jonathan Aberman’s argument in this episode is that this is the wrong question. Most people are going to use AI in their work, so authorship alone tells you nothing about value. The question that carries economic weight is whether the finished piece goes beyond what AI would have produced — what he calls originality of output, as opposed to originality of authorship.
What is the difference between originality of authorship and originality of output?
Originality of authorship is about who made something. Originality of output is about whether the result exceeds the sameness of AI-generated work. A person can author something entirely on their own and still produce nothing AI could not have produced, and a person working with AI can produce something genuinely original.
What is AI slop?
AI slop is content generated at scale without meaningful human contribution — competent, polished, and indistinguishable from what any model would produce given the same prompt. A scoring approach like Hupmapper’s gives a way to identify it: work that sits at or near zero originality intensity adds nothing above the AI baseline.
Does original content mean good content?
No. Originality and quality are separate. Something can be entirely original and still be poor work. Measuring originality tells you whether a human contributed something beyond the model’s output — it does not tell you whether the result is any good.
What is the difference between Hupmapper and Hupchecker?
Hupmapper assesses work: it scores how far a given piece of content goes beyond the generative AI baseline. Hupchecker assesses people: it measures an individual’s originality and returns an Original Intelligence Quotient. Hupmapper is the product demonstrated in this episode.
Chapters
00:00 How do you know if your team is AI-ready?
00:48 From 25 years in venture capital to founding Hupside
01:35 Demo: what Hupmapper measures, and why “was it AI?” is the wrong question
04:49 How this differs from an AI detector — no tells, no false positives
05:11 Originality of authorship vs. originality of output
07:25 Why measurable originality matters for consultants, students, and hiring
08:45 The replacement narrative, and why AI is an augment
10:20 Where to learn more
Related
Hupside: https://www.hupside.com
More demos like this: One Amazing Thing About Eikona with Nir Weingarten · One Amazing Thing About Optimizely Virtual Teammates with Kevin Li · One Amazing Thing About Scrunch with Devin Stevens
Key terms: Original Intelligence · AI slop · generative AI · AI content detection
📺 Subscribe to One Amazing Thing on YouTube — watch every demo
✉️ Get every episode by email — join The Agile Brand newsletter
Transcript
Greg Kihlström (00:00)
Having a team that understands and can implement AI-based tools and approaches is critical, but how do you know if your team is AI ready? Well, what if there was a way to assess just that?
Welcome to one amazing thing about Hupside. Today we’re talking with Jonathan Aberman, CEO at Hupside, and he’s going to be sharing one amazing thing about their platform with us today. Jonathan, welcome to the show.
Jonathan Aberman (00:30)
Hey, man, thanks for having me. No pressure. One amazing thing, you’re asking a founder to find like the one child it likes out of the 87, but I’ll do my best, man. I’ll my best. Yeah.
Greg Kihlström (00:40)
It’s a it’s always a challenge, right? It’s yeah, absolutely.
Before we dive in though, why why don’t you give a little background on yourself and your role at Hubside?
Jonathan Aberman (00:48)
Well, I’ve been in the venture capital industry for a long time, 25 years or so. I’ve seen it from different angles. And about a year ago, I saw this really amazing technology at a couple of universities and I thought it was so important to the world. I left the fund I was at and became the founder and CEO of Hupside with my scientific co-founders, another experienced co-founder. And as you’ll see, we’ve come up with something that is really, really important, we think, in the world of AI, the transformation. And frankly, it’s
pretty fricking exciting. And so what I’m gonna show everybody today is actually a demo of product that we’re gonna launch in invitation beta in August and it’ll be commercially available in September. And we’re really excited about it. I hope you will be too when I show to you.
Greg Kihlström (01:32)
Yeah, love it. Well yeah, why don’t you share your screen? Let’s let’s take a look.
Jonathan Aberman (01:35)
All right,
All right, I’m gonna run through this really fast. Here’s the situation, right? Basically what we have right now in the world is we have people creating a lot of content, we have AI creating a lot of content, and most of us are really frustrated now trying to figure out what’s actually valuable. And right now lot of the focus is around is it AI or not? And we think actually the right measurement isn’t really was it AI or not, it’s where was the value added in the piece of content? And what we have found through our science, and a lot of people know this,
AI itself is a massive similarity engine. True originality only comes from a human mind on its own or with AI together. So you need to have a way to measure that. So this is a product called HupMapper. And what HupMapper does is HupMapper takes a piece of content, written content, and actually analyzes against the baseline of how AI would generate the same content and identifies the originality, the depth of the originality specifically in that document. So you’re going to be able to come to the site
And you’ll be able to select one of these bins and tell the model what you want to evaluate. You see a social media content that you’re not sure it’s actually real. You can do that. You want to test your own advocacy piece, could do that. The second thing you do is you’re going to take up to 2,000 characters, 2,000 words, I should say, and you’re going to drop it in that box. Or you’re going to upload the sample. You’re going to hit a button. This is what you get.
you get a score, and the score is the originality intensity. The higher that score, the more originality there is in a document compared to AI. Now, a zero means it’s the same. It doesn’t mean you used AI or not. It means it is a fundamental matter. You didn’t do anything the AI couldn’t have done. You don’t want to be there, and you probably don’t want to read content like that. What you want is something that has
more and more originality. Now, the way this product works, and we’re really excited about this as well, every document has its own originality intensity score, which you’ll see rendered here. These are basically different scores. But the other cool thing is every sample has an individual fingerprint. Every one of these fingerprints is unique. They’ll always be unique. You’ll see as the score changes, the more density it gives you a resonant feeling about how much originality.
is in the document. And then if you want to test the originality of it to somebody else, or you want to post in LinkedIn and say, hey, this piece of content’s worth paying attention to, or you’re in the workforce, you’re trying to figure out if somebody’s giving you a piece of content that’s worth reviewing, this is the first time ever there’s a signal that we can use to track and see true originality content. This is open. It’s…
It’s gonna be available for consumers to use. There’ll be an enterprise version, but we’re pretty excited about it, as I’m sure you would be as well. This is pretty cool stuff.
Greg Kihlström (04:40)
So this I mean, this is interesting because I mean I feel like there are AI detector apps out there, you know, and things like that. And so, you know, that is I mean, it’s essentially like don’t cheat with AI kind of, you know, checker or something. Like this seem this seems different because you’re looking. I mean, I’m sure there’s some some in the Venn diagram there’s a little bit of overlap, maybe, but
Can you talk a little bit about, you know, how do you define originality here?
Jonathan Aberman (05:11)
Well, first of all, let me talk a little bit about the tech because I’m sure you’ve got watchers listeners to geek out on that. This is fundamentally different in that we are not tracking things to try to figure out where it comes from. We’re not evaluated against some hypothetical tells about what AI looks like. This is an absolute measurement of if AI created this content, how does this differ? It’s a completely different approach technology. And as a result, that means things like
false positives or cultural bias. It all just doesn’t matter because we’re just comparing the absolute baseline knowledge of AI. So that’s a really important thing. And then with respect to originality, this is also really important. We separate, we meaning the world right now, a lot of ways because the antagonistic way AI is being deployed and the way people react is like, what do mean you want to replace me? I’m special, which, you know, people like to be safe, like to matter, right? Visceral reaction.
Jonathan Aberman (06:10)
So the answer now is I need to know whether or this content’s human or AI. And that’s what all AI detection’s focused on. And frankly, that’s what education’s focused on right now. But there are two kinds of originality. There’s originality of authorship, and there’s originality of output. And humans are the minds that can create originality of output above the sameness of AI. It’s demonstrable. If you have originality, it’s either human on their own or a human with AI augmented together.
Otherwise it doesn’t exist. That’s a fact. The third thing is, if you take the view that originality of output is different originality, what you realize is that we as humans consume originality of output for the most part. know, the best example I can give you is if you’ve ever suffered through having a kid who wants to go to Clay Cafe and do a mug, they’ll go and do a mug. It’ll be the most trashy, little awful thing you’ve ever seen in your life.
Greg Kihlström (07:09)
Yeah.
Jonathan Aberman (07:10)
It’ll look like the other 500 on the wall, but you’ll love it because it’s your kids, because it’s original, but you’d never buy it. Right? You’d never buy it. It’s just, it’s your kids. It’s original because it’s original. But you know, at end of the day, just because it’s original doesn’t mean it’s good. Just means it’s original. So this is what I’m getting at. We needed to, we need to revolutionize how we look at AI and humans together. But if we measure it the way we want to have the world measure through Hubside.
for the first time, the true originality output that matters, which is what did you do above AI, is quantifiable. if you’re a consultant, you can show your work’s better than AI slop. If you’re trying to assess a student, if you’re trying to look at an entrance essay for a university, I mean, it goes on and on. But the key is we’re actually disabling work slop and enabling human value app.
And that’s the originality that I think fundamentally matters for the economy and for society is output plus original thought. You follow me?
Greg Kihlström (08:13)
Yeah.
Yeah. And I mean, I think it’s it it also seems realistic to me in that, you know, whether you’re a student, an employee, or or otherwise, you’re going to be using AI. There there’s expectations that you need to use AI. You know, there’s there there’s a it’s certainly a paradox a lot in the workplace and in in academia, I think, but I think it what you’re saying it makes sense because it’s again, it’s it’s assuming that there’s going to be some involvement of AI, but it’s like, what do we bring to the table?
Jonathan Aberman (08:45)
Well, exactly. What’s really funny to me is that, and this is part of the replacement story narrative that’s being promoted by the AI industry, the only way to justify the valuation is the human substitution. The math doesn’t work otherwise. The reality is that if we do this right, AI basically fades in the background like computers, like electricity, and this becomes a baseline thing that we use. 30 years ago, when I got my first IBM PC, I was a savant in my company because the first one had a computer. It’s like, oh my God, it has a computer. And then I did one, two, three, Lotus one, two, three. And they were like, you’re a genius. I mean, if you work in a company now,
Greg Kihlström (09:22)
Right. It’s magic. Yeah.
Jonathan Aberman (09:24)
you know how to use all these apps. You’re not employable. So at the end of the day, there’s going to be some things that AI is just going to be able to do really, really, really, really, really well. think a lot of things that AI does just OK.
And the things that’s really, really well are still going to need human to determine whether or not the output’s valuable. But for everything else, the special sauce of humans is still going to matter. And even if the world is like, replace, replacement, I can tell you, I bet you’re the same. I use these tools all the time and they make me more productive, but only because I use them as a sparring partner and not as a substitute. But Mike, you know, I’m about to have a book come out, Originality Dividend. Claude helped me research the book. It helped me think things through. Didn’t write it, you know?
Greg Kihlström (10:06)
Right.
Yeah.
Jonathan Aberman (10:07)
I let it write some of it and it was just embarrassingly bad. you know, everywhere that book is mine and my editors, but maybe more productive.
Jonathan Aberman (10:15)
And I think that’s where AI is going to ultimately go. It’s an augment or not a substitute.
Greg Kihlström (10:20)
Yeah, love it. Well, Jonathan, thanks so much for joining today. for those that want to learn a little bit more about the the product, where should they go?
Jonathan Aberman (10:28)
Hubsanite.com, that’s where you’ll always be able to sign up for this beta. You’ll see our other products there. There’s lots of great content. trying to make a difference. I know every startup says that, but trust me, when you’re older like me and you don’t have to go through the meld of being a startup founder again, the fact that I’m doing it, can’t. Trust
Greg Kihlström (10:50)
Right.
Jonathan Aberman (10:51)
me, if I didn’t think this important, I would still be just an investor minding my own business. That’s all I can tell you.
Greg Kihlström (10:57)
Totally agree. Love it. Well, again, I’d like to thank Jonathan Aberman, CEO at Hupside, for joining the show. You can learn more about Jonathan and Hubside by following the links in the show notes.
