The following transcript has been edited for readability. Timestamps have been removed and minor transcription errors corrected. Speaker comments and context have been preserved.

Opening Remarks

Rich Portelance

All right, everyone. My name is Rich Portelance, and welcome to the EdGate Powers Podcast. Today I'm with Brett Hodus, and I'm going to let Brett introduce himself, but our topic is AI-enhanced standards alignment and dialing in precision at scale. You know, there's a lot around this. Everybody's talking about AI; everybody within the edtech community needs to align content to standards, and there's a lot of debate on how that should happen and how AI can be of benefit. 

So, Brett happens to have an educational publishing company. He's been in this field for a long time. And so, I'm excited to have him as a guest. And, by the way, everybody, he's going to be on the panel in June. So there's a link down below to the webinar. Please sign up today. Make sure you're there; it's going to be a great panel. And as always, we're going to be exploring this topic in depth. So, Brett, without further ado, I'll let you introduce yourself. 

Brett Hodus

Thanks. I'm Brett Hodus, and I am the co-founder and CEO of Book Buddy Media. We are content developers for educational publishers. We do a lot in edtech, we are exploring a ton in AI right now, and work with about a hundred publishers around the world. 

Rich Portelance

No, that's fantastic. 

And I think you're also doing stuff in the sports media, right? You've had some background in there of how to put that into educational content. 

Brett Hodus

That's kind of the fun stuff that we do. We work with professional sports teams and help them build content that they're going to– that they're going to work with sponsors on sometimes to bring into schools, and make sure that it's pedagogically sound and also represents their brand really well. 

Rich Portelance

Good for you. That sounds really fun.

Discussion 

Rich Portelance

So, I want to start off, Brett, by talking about your perspective on how AI has changed the expectations around educational content over the past year or so. 

Brett Hodus

Yeah, I think from a publisher's perspective, it's changed in the sense that everyone feels that they're behind. Every publisher is concerned that everyone else is using AI in a way that's more sophisticated, more pedagogically sound, and better than they are. And that's probably not the truth. Everyone's probably dipping their toe in a little bit. The biggest changes I see to the content itself is, in some ways, there's a bigger gap between good content and bad content, because people are using AI in various degrees of responsibility. And so, we see some people who are just allowing the AI to do everything without a human being at the front end or at the back end. And so, you're getting this really weird robotic approach to the standards that doesn't really take into account that there's going to be a human being reading, interpreting, and learning, hopefully. And so I think there has to be a balance between human beings and AI to get that really high-level, honed-in, correct pedagogy and alignment. And I think we're seeing– because it's the wild west right now with AI– we're seeing varying degrees of success from publishers utilizing AI, and kind of stretching the limits of what a student would want to deal with in a classroom in this sort of robotic way, because even when we're teaching standards aligned content, it has to be interesting, there has to be engagement. It does not exist in a vacuum. There is a learner at the other end. 

Rich Portelance

So, I think that's fascinating, and I'm seeing the same kind of energy and uncertainty. Are you feeling the pressure to produce more content faster because of this paradigm? And what impact is that having on the quality? 

Brett Hodus

I think that we feel pressure from our clients to integrate AI into workflows successfully. So, I definitely think there's that pressure, and I do think it drives quality down. I think that AI used properly can supercharge, you know, content and make it much better, and I think we're all starting to learn some of the strategies to do that. So, I do think there's pressure. I do think some publishers are dealing with that pressure differently. I think in a challenging economic environment, it's a very tempting shortcut for people to take. And what's interesting about it is that if you take the shortcut successfully, you can have the best of both worlds as a publisher. You can reduce costs, and you can be closer to alignment. 

But I think what we need to think about as a publishing industry is when we start our engagement on any project, when we start the engagement, when we start the process of planning a workflow and planning a product, that beginning effort should be human only, because that's where we're going to get a unique perspective– a student and teacher-focused perspective– rather than a robotic “this is the standard, this is how we meet it, and then you put a human being in later”. What AI will do is give you the most obvious choices, the most obvious “what's the top hit on the internet”. And so, if everyone uses AI at the beginning, everything we have will be mediocre, and everything we have will be vanilla. And so we have to start with human beings so we have interesting concepts to talk about, to align pedagogically, to align to standards, and then have AI come in, enhance and supercharge our process, and say, “Okay, I've done this. Can you give me three more examples of that?” And it'll do that, and where are we aligning best? And where are we furthest away? So, using AI is like a brilliant friend who's sitting next to you who knows all the answers, but who is not that good with common sense. And you can't just let him roll because he'll make some weird decisions. 

Rich Portelance

Right. I like that analogy. I think it's really fitting. 

You know, so you're talking about content, and sometimes it looks engaging on the surface, right? But then it breaks down, or it doesn't actually drive learning outcomes. Do you see some content like that? 

Brett Hodus

Yeah, we see it all the time, and the reason that you see it is that there has to be, in addition to the connection and the alignment, at the soul level, a human being is going to have a light go off in their brain and realize they're talking to a robot. And I know that sounds far-fetched, but it's true. And what's happening is as AI continues to improve, human beings, especially young people, are getting better at identifying AI. Subconsciously, they'll read a sentence, and there's something in the back of their mind now– and it's new– that's saying, "Oh, this is AI." And their level of interest, engagement, etc. will immediately go off. And so, if you're going to use AI on this type of content, you have to make sure that you're connecting at the soul level with another human being, where they see themselves in the content. And I think that is the challenge, because it sort of… It sounds sort of “hippie-dippy”, but there really is– we see it with AI audio, for example. If you listen to a sentence of AI audio, it sounds incredible. It sounds like a human being. If you listen to a paragraph, there'll be a moment where you'll go, “Oh”, just for a second, and that internal mechanism in your brain all of a sudden is interpreting everything through that lens of “Oh, this is a robot. My level of engagement should go down, because it's not that engaged”. 

It's the same way when someone sends you something now, and you know it's AI. They send you this four-page thing. I got this business proposal, and it's five pages long, and it's super detailed and smart, and then you look at it closely and you're like, "This is AI. This person did this in 30 seconds”. So, they put in 30 seconds. I'll put in 30 seconds. That's the sole level of disconnect. 

Rich Portelance

That's interesting. It's an interesting reaction to, you know, what's being developed, and I think that goes back to if you are using AI, the prompts have to be specific enough, and you have to know enough to be able to really hone in and make it unique, as you're saying, otherwise it may not be an advantage at all, right? 

Brett Hodus

Yep. And also, from a publisher perspective, if there are 10 different publishers trying to align a product to NCSS (National Council for Social Studies) standards, or whatever, and they all prompt AI to give them the lessons, they're going to look very similar. We're going to have a very vanilla marketplace. So, I think an advantage for publishers will be to be really cautious about, again, conceptually, what are we building? That's a human being. That's a question for a human being. How do we supercharge it? How do we make this faster? How do we add more detail? How do we fact-check quickly? Okay, there are some tools we can use with AI, but AI should be the friend next to you, not driving the whole thing. 

Rich Portelance

So, let's kind of switch from the content development to the alignment. When you're developing content, how much does standard alignment shape the process from the beginning? 

Brett Hodus

Well, that I mean, you have to, right? You have to align, because a teacher can't use the product, right? Or they're going to be less. Well, listen, some of the best teachers will fill in gaps themselves, and they'll make it align, right? But we have to count on teachers being super busy, they don't have time, and they have to meet these standards. So, we're going to start with a path to alignment, right? We're going to build the product and build a skeleton of that product. This is the product, here's the engagement, here's the content we have to hit, here are the standards we have to hit. 

So, there's going to be a map that we're going to create and then almost build backwards. And I look at it sort of like playing golf, and instead of starting at the tee box, starting at the hole, right? And looking at it that way and saying, "Okay, well, I want to make a two-foot putt from right there at the end. And so, how do I know what shot I hit to get to the fairway and what shot to hit from there?” Sorry, golf's always on my mind, but I think that's how we should approach content alignment and standards alignment. It's the hole, it's the final thing, and so we really have to start there when we're building a product. How do we get a product that aligns so that the kids can get what they need out of it? 

Rich Portelance

And there's a difference between something that's generally aligned to something that truly meets the state-specific standards, right? And in today's environment, where Common Core has been, you know, now issued for state-to-state standards, it's more difficult than ever to make that alignment happen. How do you kind of adjust in that paradigm? 

Brett Hodus

Well, what we do as a content developer is we have state specific experts who are on the ground who really understand the standards. The political environment in the state might have something to do with it, or the funding environment, or an article in the paper, like it's very state by state, county by county, city by city, the interpretation of the standards can vary slightly. I think the standards are what they are, but I think the way that that district, and that school, and that city, and that state apply the standards, how rigorous they are in terms of how closely aligned to those standards need to be, what are we giving up on the interest, content, the selections, the choices we're making, and the student engagement. How do we marry all of that up to get to that final point? 

So, I think it's really a state-by-state thing, and it's making sure you have the right expert who doesn't just read the standards like an AI would, but someone who understands the environment that those standards live in. 

Rich Portelance

And I know that's been a point that EdGate has, you know, been very firm on and in making sure that they have experts at the beginning. And we're very hesitant to input AI into some of their interworkings, you know, with ExACT, which is an alignment-at-scale kind of product. They were cautious about doing that because of that loss of personalization. I think they found some ways in which to implement components of AI so they can, you know, do more with less, but without losing the human element. 

So, I'm curious about your perspective on alignment at scale, because, as you just described, you have experts at the state level who understand how to write the content, you have to be able to align using those experts. But now as a publisher who wants to sell to 50 different states, how do you marry that all up, and what do you see as the trend in the market? 

Brett Hodus

So, it's interesting. 

So, I think that when you're a publisher and you're looking at scale production across 30, 40, 50 states, whatever it might be, you really need that expert at the state level, because otherwise, think about this. If we have AI, I'm an expert at anything. I'm a plumber, I'm an electrician, I’m a business analyst, I'm a banker, I can do anything with AI. But, like, should I? That's the question. And so, you could theoretically use AI and say, "Yeah, I'm an expert in state standards in Kentucky now, because AI will tell me what they are." But if you're not on the ground, if you don't have that nuance, if you don't understand the environment, the way teachers are interpreting it in that state, you are going to miss. And you're going to miss in the same way everyone else misses, right? 

So if you don't bring in that expertise, again, at that state level– and so, how do we use AI to save money on scale content alignment? We use it at the sub– we save money at the subtask level. It's not the task level. It's maybe, you know, maybe the AI is getting us the standards, and maybe it's also culling through Reddit feeds, looking at different teacher websites, and saying, “What are some of the issues teachers are having?” And maybe it's assembling all that stuff so that our expert has more information. Our expert has more context. They get supercharged in a sense. Maybe the AI is doing the layout of that correlation. Maybe they're writing it on Word, and it's like, “Hey, lay this out so it looks clean and put it in a chart”. This is where we should use AI at the subtask level. But we should not assume that we're all of a sudden experts in every single thing in the world because this AI is standing next to us and knows everything. It knows everything. It's brilliant, but it's super dumb when it comes to common sense, and it'll do weird things. And if you give it the wheel, it's going to drive you off a cliff in a sense. But if you put it in the driver's seat next to you and ask it questions and have it help you navigate, you're going to get where you need to go. 

Rich Portelance

It’s the Michael Scott paradigm of sitting in the car and driving into the pond, right? “The navigator told me to go there!” 

So, let's get into districts and, you know, when your content gets to that level, what feedback do you hear from educators or administrators at this point? 

Brett Hodus

We don't hear as much directly because, you know, we work with our publishers, and they set off sail. And so what we hear is through them, and generally our publishers tell us everything's great all the time. So, we don't get a lot of direct feedback on that. I know that's not a good answer, but they definitely are coloring their products as being the answer, right? So, we get a skewed perspective in a negative way there. We don't get a lot of direct interaction at the end after we build a product with the person using the product. 

What we see are general trends that we see through our publishers. We see what products they're building or not building, and then we infer what the market said about the product they built or didn't build. What I can tell you on the standard-specific front, that's where we do get very specific feedback, where a publisher will come out with a product, and that's why we're often hired to audit a product, because they may have built it to align to standards maybe three, four years ago. It's not working in the market for whatever reason. We'll come in and say, "Hey, we're bringing in an expert. We're supercharging it with AI, and here's where we think you're missing." And so that's kind of more our role is through that kind of second, you know, that game of telephone. 

Rich Portelance

That's interesting. So, as you're sitting in that passenger seat in a sense, you know, helping to drive and navigate toward success, which is standards-aligned content that actually works in a classroom, do you see things that tend to break when content isn't aligned closely enough to standards or curriculum? 

Brett Hodus

Yes. So, the top things we see is, number one, the first thing, and it's the hardest one to measure, but we definitely hear it anecdotally, we see it in the student test scores. Kids aren't getting stupider; they're getting less interested. And so what happens is you get a kid who just doesn't try. And effort is the number one indicator of success, right? How hard are you trying? And so, I think that's the first thing. 

I think the second thing we see is that we do not see alignment within a district, or within teachers, even in the same department, on how to use content. Because if it doesn't align, there's always that great teacher who's making it work, and these other teachers who are saying, “How are you doing that? I don't have time to do that”. “Oh, well, I wrote this supplemental, and I did this other thing”, and, you know, they bought it, they shouldn't have to do that. It should be done for them. And so I think that's what you get. You'll get a couple of classrooms that are using the product successfully because you get a superhero teacher, and then you get four or five others that are saying this thing's not working. 

But I do think the biggest thing, and again it's the hardest thing to measure, is that student effort and engagement. It's just not the same when the content is not focused on their interest. 

Rich Portelance

Okay. That makes a lot of sense. Are you seeing your publishers, the people that you work with, struggling with understanding and reacting to state-level regulations that are cropping up? Whether it's third-party reviews or other things that are more specific, rather than just, “I'm selling my product in here now”, you have to jump through five hoops before you sell your product? 

Brett Hodus

Definitely. And I'll give you the example with this California adoption, right? And so, you know, and I think part of this actually is rooted in AI. I think a lot of the information from states is now being developed with AI, and so the information is getting longer. The websites are having more information, and yet it's impossible to find anything. Sometimes the information will be contradictory. 

Rich Portelance

Yep. 

Brett Hodus

Sometimes the information will be incomplete– oftentimes incomplete. And so, I see again, one of the impacts of AI, because it impacts everyone, everywhere, is that at the state level, I think they're developing some of their standards and some of the rigor they want us all to go through using AI to an extent. And I think that is going to be counterproductive when we're talking about something as nuanced and tricky as alignment. And so I think they try to do a one-size-fits-all. It's missing details, and it's contradictory. They're fielding hundreds, and hundreds, and hundreds of questions. And I'm not calling out California, this is every state when they're putting out their information. But that's what I'm seeing. I'm seeing a confusing landscape where it's very hard for me to know what is my target? What am I trying to hit? Because I've been through 50 pages of this website and 12 different links, and it's hard to find the information. 

Rich Portelance

Do you have to find an expert? Or, you know, how do you navigate that environment?

Brett Hodus

Yeah, that's usually what we do. We usually have a combination of, you know, internal teams that are used to doing correlation projects, that are used to doing alignment, and then mixed with that state-level expert to go in there and say, “Oh no, ignore that part. That was from two years ago. They didn't remove it”. They know that, but we don't know that, right? There's stuff like that. This is an imperfect– the materials they're giving us and any publishers who're listening to this who've been through any adoption know exactly what I'm talking about. You go on that website and you're like, “Gosh, it says this here, but then I click on the link and it's expired. Or I click on the link and they're not releasing it till April. Or this says 2024, it's 2027. Does that still apply?” And so, when you get someone on the ground level who is an expert in that state, they'll say, "Oh, no. Those are the old standards. They're updating it next week”. They just know stuff about what's happening that's not necessarily listed on the site. 

Rich Portelance

Okay. 

Brett Hodus

And if it's not listed on the site, AI doesn't know it. 

Rich Portelance

Got it. Yeah. I’ve been, you know, talking to EdGate and, you know, my sense is that over time, there has to be an end-to-end solution where standards creation, standards alignment, can be connected to the, you know, database of standards. This is so folks like yourselves can find where the heck these are because, you know, for EdGate, they've had to develop expertise in search and “recovery missions” just to get standards into the database. That shouldn't be so difficult. 

How do you kind of automate that process and then the alignment process, and have more control over it, so that when the content does come out, everything can be found? So I see there's an opportunity there. I don't know how it's going to get tackled, but it's very desperate right now and extremely confusing as you just pointed out. 

Brett Hodus

Right. Well, yeah. And I think what people tend to think is, “Well, the AI will keep improving and that'll solve it”. The problem is that the AI can only learn from what's on the internet, and it becomes a loop: if bad information is on the internet, that's what it'll learn. And we see that the information– again, anyone who's been on one of these state websites for adoption– you see that it's 85-90% accurate. It's missing some– it's not 100%, and the AI is learning everything it's learning from accessing all this flawed information, and it becomes this loop that if you don't get on the ground and really dig in, it gets worse. 

And the other part of it is human nature. There's a real tendency to be lazy about different things. We're just human beings, and we want to do other stuff besides work, so we're like, “Hey, I can spend three hours digging in and writing a report about what we're missing here. Or, I could just ask AI and hope it's right, and sometimes it is. So, let me just do that and go home early”. And so, that's happening everywhere. It's tough. I mean, so, we're going to get weird results for a while until people really get an understanding of what these tools can and can't do.

Rich Portelance

So, you know, I imagine– you've pointed out some of the issues with AI. Where is it actually useful in your content creation workflow? It doesn't have to be creating the content, but do you have a workflow that says, "Hey, there are some advantages here"? 

Brett Hodus

Yeah, absolutely. And we're using it every day for subtasks. And so that's why I keep coming back to that same thing is, you know, like I'll just give you a couple of examples. We produce audio, so on the audiobook production, the first round of review we do is with AI now. If we do an audiobook or we do a textbook, AI reviews it first, then we make our revisions, we make our changes, and then the human being goes back and is the final review on it. So, that's been helpful. We used to have two different human beings look at it. So, there's a bit of cost savings. It's not a significant cost savings, but it's a cost savings since we pass that onto the customers. 

When we're talking about building, like, library books, like fiction content, something like that, or non-fiction content about the weather, or rocks, or earth science, or whatever that might be. What we used to do is have AI help with the outline. Bad idea. Because like I said earlier, then you get this really boring book about the most obvious thing, whatever's already been written about a million times. So now we have human beings do that first outline, but then we have AI do an enhanced version of it. So, we'll take that outline and we'll say, "Hey, give us notes about what images should be where”, “give us notes about other examples that might be better if we missed one". And so we try to have AI make our book better, make our process better, and that might be a step that we would have had with sort of an editor at that stage. So, we'll have AI do that. Then the writer will come do their thing, the layout person has this awesome queued up thing, “Oh, I should find a photo of a man in a hat near a cliff. And I can do that”. And we'll even have AI put in the prompts so that they could be copied and pasted into Getty, so that the designer can just go, "Show me this. Show me this." 

So these are the little things we do, but it's still a human being designing the book. It's still a human being writing the book. It's just, we're trying to do little things to make our jobs easier, and to make us a little bit smarter and have information at our fingertips. But again, we're not letting AI drive that car. 

Rich Portelance

So, you're definitely putting human judgment into the loop to make sure things come out, you know, with a human feel. 

Brett Hodus

And what we're seeing on costs is that we're able to reduce costs depending on the project between 3% and 10%. Not 30% or 40%. If you're getting a 30% or 40% cost reduction, you're doing it wrong. You are letting AI drive you a weird product. It will not sell. So that's what I'm seeing on the ground level. Now, again, it may change, etc., but it requires a real, thoughtful workflow to make sure that everything AI does has a person at the front and a person at the back– It's an AI sandwich. 

Rich Portelance

I like the idea of that workflow. You know, I do believe that AI in the background of companies helping to create efficiencies driven by human input can, you know, help companies scale, help companies improve, but it's not replacing, it's augmenting. 

Brett Hodus

Exactly. And there's pressure, I mean, there's definitely pressure on CEOs, especially at the large companies, to, “Hey! Reduce cost by 20%. We have AI now”. And it's going to be interesting what educational products look like in the next three, four, five years, as these products that are just now being developed, and maybe people are leaning too far into AI, what those look like, and how it impacts outcomes. And I think there'll be a real windfall of business for companies that are able to thread the needle and use AI to reduce costs, make the product better, rather than drastically reduce costs and not worry about what the product looks like. 

Rich Portelance

So, considering that paradigm, what you're doing, and how you're leveraging AI, you know, in that process, when content is done well, both engaging and aligned, what does that actually look like?

Brett Hodus

Great question. 

It looks like the best products that you use, right? It looks like a student who's engaged, a teacher who's checking off every box because she's getting the standards hit, one by one. More importantly, it's a product built on understanding how much time she has or he has, right? So, I think a lot of times, you know, you can have the best product in the world, but if I have 40 minutes to use it once a week, and it's built to be used three hours, twice a week, it won't work. 

And so, also aligning that product– the interest for the student, number one. The standards have to be hit for the teacher, 1A. And then maybe 1B is that it's got to be implementable; you have to be able to do it. I don't know if implementable is a word, but you have to be able to implement it in the time you have for it. And then, is there a chance for students who fall behind to get lifted up? Is there, you know, some type of supplement for tier two or tier three? Is there some kind of teacher support material to make sure that the teacher can hit all the students who are getting it and the ones who aren't getting it? 

Rich Portelance

Okay. And in that paradigm, are there certain signals that tell you content will succeed in a real, clear classroom? Not just theoretically. I mean, you kind of touched on it, I think, or alluded to it, but are there clear things when you say– when that's going out the door, “That's going to be successful?” 

Brett Hodus

Yeah. I think, you know, this is just an interesting thing that I believe. I think the people who work on the product, the more fun they're having building it, the more they want to talk about it, like I see this on my team with certain products. They just want to talk about a product, they're excited about it. I think, to me, that's an indicator. I don't know how much that helps everyone else out there, but from my seat, I see some projects where I feel like, “Oh, the client's asking us to do this thing and we're going to do it, but I don't know”. And then there's other products, “Well, wait till you see the artwork. Wait till you see how great these lessons are. Wait till you see how easy this is to implement, and you can feel that passion from the creators”. And I think, again, that can't be replicated. Like AI feels the same way about everything it produces: neutral. 

Rich Portelance

Yeah. 

Brett Hodus

Right? But real people, they know when they're making something great. 

Rich Portelance

Well, and it goes back to your earlier comment about, you know, engagement with students. You know, if they're bored, they're not engaged. And if they're not engaged, they're not learning. And it's, you know, incumbent upon the producers of the content and the teacher to make it engaging so that the student can, you know, can learn. 

Brett Hodus

Yeah, and of course, we look at, like, education over the years, and you think about, like, you know, as we price things, like if we price a library book, for example, to a library publisher, and then we take that same book, and we price it to a trade publisher, what's the difference? Well, the trade publisher is going to say, “I need better artwork. I need better this, I need better that, because I have a consumer who has to want to read this”. So do we! We have the same thing in the educational market, but a lot of times, educational publishers forget that we need to have the consumer, which is the kid, the learner, to want to engage with the content. And so, even if it's a textbook, even if it's, you know, hardcore curriculum, I think that that paradigm is being exposed with AI, because you could see in the trade market and in the retail market, they are much more resistant to AI. You're not going to see a best-selling book with an AI cover right now. It's not going to happen, right? But in the educational market, you might see that. Why do we lower our standards for the kids who have to learn and allow that to be sort of normalized, when we would never expect anyone to go into Barnes & Noble, or Amazon.com, or Costco and buy a product not produced by a human being? 

So, again, background subtasks; make it better, supercharge it, but at the front end, you know, we've got to connect at the sole level to the user. 

Rich Portelance

All right. And you know, I guess talking about some market signals, because you've just touched on that, are you seeing buyers place more emphasis on alignment today than they did a few years ago or less? 

Brett Hodus

I think if anything more, if anything more, I think it's– I think it's definitely… And I think because the outcomes aren't great, we're seeing scores that aren't good. And so, we're really trying to hone in on getting a consistent “how do we hit everybody, and how do we make sure everybody is up to snuff and you have to have standards?” That's just the only way to do it; otherwise, you're relying too much on the teacher-by-teacher, and they're not all built the same. There are a lot of super teachers, and there are a lot of good teachers. And if you have a super teacher, it doesn't even matter what the curriculum is; they're going to hit it out of the park anyway. But there are not as many of them as there are just regular teachers. Super teachers are like a super anything, right? There’s like, it's 10%, it's 5%. And there are a lot of kids out there that are lucky enough to have a super teacher, but the rest of them have really good teachers who have lives and do other things, and they need us to help. They need us to create content that is aligned, interesting, engaging, and that they can teach in the time they have to teach. 

Rich Portelance

So, is there a feedback loop from those– or, I guess, let me rephrase. I'm curious about subjects, and are you seeing more success in certain subjects that you're producing than others, because of some of the challenges with alignment? 

Brett Hodus

I mean, I think math is an easy one to align to generally. Science… NGSS (Next Generation Science Standards), NCSS, like there's those are sort of easy. NGSS, they're sort of easy to align to. When you get into English language arts, the Science of Reading is tricky. There's some weird stuff going on there. And there's a political environment where some people are on this side and some people are on this side, depending on their stance on, you know, should we be doing 100% decodability, or should we be doing 90% decodability? So there's some disagreement in the market. So I think ELA is tricky because of that environment; I think people really focus on it. 

But yeah, other than that, I don't think there's any outlier. That would be the outlier, probably, is ELA. 

Rich Portelance

And do you have databases of information that say, “Hey, in this particular state, here are the rules, this is what we're going to follow”, versus this state, so please know the difference?

Brett Hodus

Definitely, definitely. Just as an example, is like, you know, there's certain states that are Science of Reading states, and you just have to be really be cautious with that decodability, and you have to get aligned with a certain, you know, UFLI (University of Florida Literacy Institute) as an example, they have, like, the University of Florida, they you have to be aligned with UFLI in certain states, and then other states that not as much. 

And then the other place we see this, where, like, the environment of the state is hugely impactful, is at the branding level. For example, social emotional learning. You used to be able to market and sell social-emotional learning in 50 states. And now, if you use the word social emotional learning in certain states, not only will you not be able to sell, you won't even be allowed to sell. Forget about whether it will sell, you'll be pulled off lists, like, you can't use that word anymore, you know, in states like Texas and in some other places. And so, you know, just knowing that, having an understanding of that, and having partners that understand that is important to publishers because you put out the best product, you label it with one of these “dirty words” in a certain state, and you may find that your sales are zero. 

Rich Portelance

Which is an interesting paradigm and, you know, kind of environment that we live in. 

Brett Hodus

For sure it is. 

Rich Portelance

It makes it challenging. 

What do you think the market is still underestimating about content quality and alignment? 

Brett Hodus

I don't think it's underestimating anything. I think everybody kind of shares an anxiety about AI production. From teachers, to students, to publishers, to, I'm sure, state departments of education. I think there's a shared anxiety and a shared worry about what is the right type of product, how aligned, and how much AI. I think it's a weird moment. So, I think there's a shared anxiety from all parties. 

Rich Portelance

So, I'm going to give you a microphone, and you're, let's say at ASU+GSV (Arizona State University + Global Silicon Valley), and you're in front of a couple hundred startups. Every single one of them has AI in its title somehow, and they're all coming into the market thinking they're going to take on the world. What are you saying to them? 

Brett Hodus

Oh man, this is a good one. 

Two things. The first thing I'm saying to them, and I actually did this exercise a couple of weeks ago, and I went through, like, a 100 edtech startups. I just looked at websites and tried to figure out what they're doing. One of the things I would say is that your AI tool is not that interesting. There are a lot of AI tools that are doing a lot of similar things. And the big, big, big companies, Google, Amazon, whatever, they're going to start building AI– and they already are–into their tools, and they're going to make a lot of what you do not important, because a publisher is going to be able to do that through some other tool. You can't– like it might have been innovative three years ago to say, “We could write custom lessons for every student”. It's like, well, everyone can do that, I could do it with ChatGPT. And so, it's not that special. So that would be my worry for them. 

The advice I would give them is to have print companions. Make sure you have printed content that goes along with your edtech product so that you have something tangible, interesting, and branded to sell, and it's not just kids in front of computers. We saw that with the Los Angeles Unified School District in the news yesterday or the day before, with screen limits for students, etc. This is going to continue to happen. And so, if you're an edtech company and you have a great solution, number one, tell me how it's different from ChatGPT. Tell me how a teacher can't do exactly what your thing does herself, that's number one. And number two, how do you get some print components so that your product is tangible and is not trapped in the computer that people are trying to use a little bit less of? So that would be my two things. 

Rich Portelance

Print. What's that? 

Brett Hodus

No, exactly. But I think, you know, there's going to be this push for nostalgia, right? We're already seeing it. Anyone who has a teenager and I have two of them, getting their head out of that phone and to look up and be a part of the world is a challenge, right? It's a hard thing. And so if we keep building… The way they're entertained is on that phone. 

Rich Portelance

Yeah. 

Brett Hodus

If we also build their education inside of that phone, how do we get them out of the phone to, like, to interact and do stuff? And, like, be a productive member of a team at work, or whatever else? Like, you have to get your head out of that phone. And if all of the educational technology we're building puts your head back in that phone, we're missing it. 

Rich Portelance

You know, I thought about that. My wife's a teacher, and I know her school has, you know, the kids put their cell phones into a pouch now. They're not letting, you know, the cell phones be an active part of the daily rigor of the classroom. It's going to change. And, you know, as you know, you and I have similar backgrounds. I was a graphic designer coming out of college, and I know you went to NYU (New York University), and you were creative. You know, we think in different terms, and back then, it was in print terms mostly. But as, you know, the digital world kind of expanded, you looked at, well, you know, there's complementary components there, but then going all digital all the time you know has its limitations. So I am glad to see it swing back and I'm glad to see some people think in terms of a physical product, because it has a different… You know, that the tactile nature of it really engages people in a different way. 

Brett Hodus

Well, and this is, again, not to throw away, like, AI is the most amazing thing. I use it every single day, and I encourage everyone to use it every single day. I think it's awesome. I think it's ultimately going to help education. But there is no silver bullet. You can't just give it the wheel and say, "Go take us home. Go teach all the kids, and oh, we don't even need teachers. We'll have a robot up front that'll go like…” You can't do it. That's just not going to work. It's not going to work for kids, they're not going to learn that way, they're not going to be engaged that way."

Closing Thoughts 

Rich Portelance

All right. So, I have one final question. I think it's a culmination of what we just talked about. What's the one thing content creators should be doing differently right now? 

Brett Hodus

They should be identifying every task and subtask that they do, and making a gigantic list, and figuring out what each task costs for a human being to do, what the AI compromise is, and whether it is worth it on a task-by-task level. 

Rich Portelance

That's a fascinating exercise. I'd love to see some people actually do that, Brett. 

Brett Hodus

We're doing it. We're doing it at Book Buddy Media. We have a list right now of 314 tasks and counting of all the things we do, and we're trying to put a human cost to it and an AI cost to it, and then a note about, is this something AI should do? And we're finding in most cases, the answer is no. In some cases, it's like, we found one, we did it. This is something AI can do, and let's figure out how to integrate that. And so, you can't look at the big picture. You have to look at the small picture, task by task, so that you don't give it the wheel, but it helps you drive better.

Rich Portelance

Oh, that's fantastic. 

So, you know, again, I'm Rich Portelance, this is the EdGate Powers Podcast. We're leading up to an exciting webinar here in June on AI-Enhanced Standards Alignment: Dialing in Precision at scale. I'm with Brett Hodus, the founder of Book Buddy Media, and you know, thank you, Brett. I'm gonna let you have the last word today, and I thank everybody for joining us. Brett? 

Brett Hodus

Well, thanks so much for having me here, Rich. I appreciate it. Good to talk to you. It's a really exciting time to be thinking about AI and figuring out how to use it in the content we develop. So, we'll see how it all unfolds. 

Rich Portelance

Awesome. Thank you, Brett. I appreciate it. 

Brett Hodus

Thank you