AI Real Talk: The Curiosity Connection

Pete Hall: Welcome to Artificial Intelligence Real Talk.

I'm Jethro Jones, and I did some interviews at ISTE, which you can hear on my other podcasts as well.

And this one is with Mike Lawrence, where we're talking about artificial intelligence, so I thought it'd be perfect for this show.

Hope you enjoy, and thanks for listening to Artificial Intelligence Real Talk.

Jethro D. Jones: so what, what would you like to talk about today?

Mike Lawrence: Well,

I'd love talk about AI, uh, trust and transparency that we've deployed in my district, if

Jethro D. Jones: Yes …of interest to you.

Okay, that sounds good.

So, uh-

Mike Lawrence: sticker too.

I

Jethro D. Jones: I have a sticker too on it.

Oh, good.

Thank

Mike Lawrence: our slogan and our website, so I had them printed from our print shop before I came.

Jethro D. Jones: To,

Mike Lawrence: yeah, start a conversation, not an accusation.

It's, uh, an approach we've taken to deploying AI in the

Jethro D. Jones: Okay.

Very good.

like it.

Um, all right.

As we're talking, just hold that, uh- microphone up close to your n- mouth.

That'll be good.

Um, all right.

Mike Lawrence, ISTE.

Uh, Mike, welcome to the show.

Glad to have you here.

Our show is called Artificial Intelligence Real Talk, so this is, uh, timely that you are here.

Um, tell us about your approach to, uh, AI policy and transparency in your district.

Mike Lawrence: Absolutely.

So I started at the district, um, over two years ago, uh, and, uh, we had a guidance policy that was being developed, and our primary focus at the time was HI plus AI.

The idea that, uh, AI is dumb as a post if you don't involve the human.

And so, uh, HI, human intelligence, and AI work together.

And so we initially came up with the formula of HI plus AI, and then ultimately decided it really is HI plus AI plus HI plus A… It's a back

an iterative process.

Um, we took that out and shared it with our parent community and our, and our, um, uh, additional folks in the broader community, teachers, administrators, students.

We asked for student voices to come into the process.

And through all of those voices, we realized we really need to take a different approach.

You, you, you don't wanna s- ban things.

You don't wanna block things.

We needed to find a way to be proactive and welcoming in the that was So we started working around this idea of AI badges, and we initially had one
version of it that, uh, that you could share, "Hey, this was my idea as a human, and then I modified it with AI." Put a badge on it that says that.

Uh, and these, by the way, these were John Spike, uh, his idea.

Uh, he's an educator out of Wisconsin, and I saw it online.

I was like: Oh, that's what we need.

And then, you know, there's another badge that says, "This was an AI-started thing, and then I modified it as a human." And we used that for a few months, and we realized it was not asking the right question.

What we needed to know was, are you using AI?

And if so, how much?

And, uh, by building it out as something that's, that's transparent, that I'm sharing with you out of my own choice, um, I can then Uh, control that conversation and it removes some of the taboo.

So we then redesigned those badges

to to be, uh, built around the idea of is it AI?

Is it human intelligence?

And if, if it's AI, is it 20%?

Is it 40%?

Is it 60%?

And we built a series of badges for students and staff to put up on their, their documents- when they build things.

Jethro D. Jones: So let me ask you a question.

Why does it matter if they're using AI or not?

Mike Lawrence: A great question.

So, uh, and I'll go back to policy on that.

Uh, when I first got there, they said, "Mike, we need to update our policy 'cause AI's a thing and our policies are years old." And I read policies, the academic honesty policy, the student use of technology, and the staff use of technology.

And in reading it, I didn't see a need to change it because it talks about be honest about your use of work.

Don't others' work as your own, and I think that includes AI.

Did it call it out when it was written in 2009?

course not.

But I argued that we didn't really need to rewrite that.

So I compromised, and I said, "How about I create an AI addendum?" Mm. initially created an addendum where it called out specific lines from the, the policy and said, "Here's what that means in an age of AI."

And it didn't change the policy.

policy was still the same.

Um, you know, don't use AI to harass another student.

Uh, don't, don't claim work as, as your own when it's not your own.

Uh, you don't use personal, you know, pers- you know, work devices or school devices for personal activities, which could be AI.

And so we called out specifically all that, what that looks like, integrity and responsibility and ownership.

And so that was the reason that sort of led to, oh, we should probably make it clearer when AI is being used and when it is not.

And the second thought I had around that was teachers, we generally assume that we can ask the students any time, "Did you…

Is this work your own?

Did you use AI?" But we don't ever feel a compulsion to share with the students-

Jethro D. Jones: the students- That's what I was just gonna bring up … whether

Mike Lawrence: we use AI.

Jethro D. Jones: So one of

Mike Lawrence: I saw that as

Jethro D. Jones: Exactly.

That's what I was just gonna say.

I find that very hypocritical we demand to know from students whether they're using it or not, but never feel the need to disclose that ourselves.

We're like, oh, it's this whole, uh, good for me, but not for thee kind of a thing.

That just drives me bonkers because that's not, that's not valuable to our kids, nor is it setting a good example

Yeah.

And so like, what are we-- We're also just saying, "We're in charge, we're in control, and you just have to do whatever we say," which is not helpful to kids either.

Mike Lawrence: Right.

Yeah, and there's students that are t- they're very upset, I think rightfully Mm-hmm … when they can clearly see that feedback is written by AI.

And then the teacher is like, "Yeah, but you can't Yeah.

So this was rolled out, the, the transparency badges, we rolled them out for staff.

We started with staff.

We started with administrators.

My superintendent includes a badge whenever she does a presentation.

Uh, our communications officer, when he sends out the Monday message with all the updates of here's what's going on in the district, he uses AI to help streamline his language to make sure it looks good and that it's voice and the transitions are clean.

Uh, he's an excellent writer himself, but he's using it to, you know, streamline and make sure he didn't make any dumb mistakes- and improve, and he puts the badge on those emails.

Um, and so then we said, "Okay, staff, keep using this and, and show this to the students and model it yourself." Then when you ask the students, "Did you use AI?"

There's less of a taboo because you've already been open about your

And the phrase we use, our slogan is, "To start a conversation, not an accusation." And we try to live that and embed that with these badges and our own policies, and we have another tool as well that we use to try and take it even further.

Uh, but that's the core principle.

Mm. the students see the teachers being upfront about it, they're more willing to be upfront themselves.

Uh, and we hear from the students that it made, it made them feel relieved- that they knew that the teacher was okay with this type of use and this amount of use, uh, but maybe not this And it gave them clarity that they were lacking.

Jethro D. Jones: I-- Well, I think the clarity piece is huge because if you know what the standard is, then it's a lot easier for you to, to follow that standard.

But if you have no idea- then you're like, "Well, I better not say anything 'cause I don't know- if this is okay or not." Yeah.

And that, that's really important.

Mike Lawrence: Yeah, and the, the blanket bland- ba- bans don't work.

We knew that from, you know, the "Just Say No" campaign '80s.

Yeah.

We, we, we… I heard growing up, "Just say no to drugs," but I was never told what I could say yes to,

you So why, if we're saying, "Just say no to AI," well, what e- what can I say yes to?

What, how can I use it?

And how do we define AI, by the way?

'Cause does that mean that spellcheck, I can't use

that AI?

C- uh, can I not use my GPS 'cause it's calculating where I Yeah.

Can I not use Grammarly to help with things?

At w- where do we draw the line if we're gonna just put a blanket ban?

aspect

Jethro D. Jones: and the other thing is, where do we draw the line on disclosure when- we know that we're all using it all the time- for all kinds of things, even if it's not, you know, generative AI.

I used ChatGPT to rewrite this is different than I used Grammarly, but the technology there is basically the same.

The computer is saying, "This is a better way to phrase this." Yeah.

Mike Lawrence: Yeah.

So,

Jethro D. Jones: Uh, you know, I, I think about people in industry and, and a lot of things that they're doing, uh, not in all roles, but in some, is if you're not using AI, then, like, you might not have a job.

Because it is, it is that important- Yeah

to, to use it.

And so to think about disclosing it, like, that's not-- That's kinda crazy because

…if you said, "I did this all myself and not AI," they'd be like, "Why'd you spend so much time doing that?" "We just need this done." Like-

Mike Lawrence: At some point, it will become the disadvantage that

Jethro D. Jones: Yes, yeah

So how do you, how do you see in the future educating kids in a way that AI is a, a f- an AI first type of approach?

Mike Lawrence: Well, it's a tool, right?

And I think that every tool that is introduced has a backlash and a banning phase.

Um, and I think we would be foolish and, uh, unfair to our students if we didn't prepare them for a future that will have AI.

And it'll have AI that's vastly better than the AI we have today.

Every year, every minute, it gets better.

So I feel that it's just a natural progression for us.

There's always a backlash to it, and then we move forward, and then we become AI-first, AI-positive on it.

Um, the comparisons I've heard that I, that, that make me chuckle is, you know, you're running a warehouse, and someone invents the forklift.

Are you gonna go, "I'm not using that"?

Of course you're gonna use it.

Of course you're gonna figure out a way, because then you can leverage the people you employ much more efficiently and faster and get more shipments in.

You're not laying people off.

They're just able to be more productive.

They're able to move things in that warehouse much faster, and you can then focus their time on the things that matter, like increasing the shipping and increasing the, the, the in and out of that work.

And so we look at that for, um, our work as let's free the teacher of some of the drudgery and give them the opportunity to provide high-quality, real-time feedback to students with their human i- uh,

intelligence, and have the AI do the sort of rote repetition and And I think the same is true of students.

Let's get them using their creativity, their empathy, developing their skills as a writer, as a, an artist, um, by leveraging AI to just check some of the dumb stuff that, that could lose points and, and, and hurt their work.

So, so that's one of the areas that we feel about it for the future is that I think it's…

Even the badges, I think we're gonna look back at as

quaint.

Yeah, The badges, I realize, are a transitional stage.

Yeah.

It's only gonna be useful for us for a few years because, uh, after a f- yeah.

'Cause after a while, you won't be able to identify how much AI is being used and where.

So the functional aspect of the badges of being able to say 20% or 40%, even now it's subjective, uh, and we know that.

But at some point it's gonna become useless.

Yeah.

So, so this is just a way to get our mindset around, uh, being open about it, to get those conversations started.

And then once we get to a point where it's natural, then- I don't think we have to keep using the badges.

Jethro D. Jones: Yeah.

So about that, one of the things that I appreciate about that is that it is a mindset thing and that it is a human behavior thing- as well.

And so you can't just be like, you know, "Let's, let's just, you know,

uh, add

these AI tools and then everybody's gonna use them and we're gonna be fine."

You have to change the human behavior as well, and that's, and that's what you're doing.

And so I appreciate that depth that you gave to that answer to say this is about changing what they're doing.

It is transitional, and eventually it's not going to be that way.

And I think about, uh, as it relates to AI, this transformative technology that we're experiencing, that it's a lot like what electricity was back in the early days.

You know?

Like factories, like electricity came, and it's not like all of a sudden the factory was like, "Boom, we're done. Now we're all electric." Um, they still used whatever systems they were using to make it work.

But then once the, the infrastructure got there to where they had electricity running to the factory and they had lights and they had machines and all this kind of stuff, then they started using that stuff and, and making it, uh, work for them.

And so to me, there's still part of this where there's this infrastructure that doesn't exist.

Like, uh, SDA SCD, uh, is talking about this thing that they're doing called Stretch AI.

Yeah.

Which is a, um, a, a chatbot basically trained on their stuff.

But the problem is a chatbot is like, that's like yesterday, and we've already moved on and we're using agents and things now.

Mike Lawrence: Yeah.

And,

so

Jethro D. Jones: that, idea of having a chatbot that can do that, that's fine, but that's not getting the infrastructure in place.

To me, a better way to do that would be to say, "Here are our APIs.

Members can access them So your agent can go find- Yeah … the information that you need.

And that kind of agentic type stuff that our kids are gonna have access to, that's, that's the infrastructure that's going to exist, where they each have their own little agent where
they say- "Help me with this kind of stuff." And, and that to me is really what, what future looks like, and, and we need to figure out what that infrastructure looks like for everybody.

Mike Lawrence: Yeah.

And I think in terms of Stretch AI, I think it's a necessary step, and perhaps it's a year or two late.

I think development cycles being they Even with the resources of SDA, SCD, they're still a nonprofit.

In the scope of generative, uh, AI tools, they aren't gonna have the resources to throw at the Yeah … phase.

But I do see it as a, as a fundamental step that then could turn into an agent.

Imagine if you had essentially Stretch AI became a port they, they could pop into and say, "Okay, what does the recent research say?

What does A- uh, SDA, SCD say about, you know, screen time?

What is it gonna say about this?" And all of that research that they've conducted is now accessible, um, beyond a chatbot.

I think a chatbot is, as you've mentioned, sort of yesterday's news.

But I think once they turn it into an agent of some kind that your agents can interact with, my agents with- Mm-hmm

and we can collectively build knowledge, um, I think it would be hard to do that without having the foundational Stretch AI piece.

even if it's not as immediately functional as we all would like, it's better than not having it, and it's a valuable resource, I think, for us moving forward.

other nonprofits would do well to pool their knowledge into some tools.

Set up those APIs it available to agents.

Jethro D. Jones: Yeah.

And then, you know, that way you can have your stuff out there, but you don't have the overhead of, of, uh, training an LLM, which even if you're fine-tuning is still an expensive endeavor, you know?

Yeah.

Mike Lawrence: But you're also s- uh, putting a wall around it, so it's not grabbing junk data from Reddit gonna grab researched, uh, you know, vetted articles from your authority, whatever that Yeah.

Whether it's ASCD or others.

Um, so yeah, I think- Yeah … I a valid

Jethro D. Jones: Um, are you guys doing anything with agents right now?

Mike Lawrence: Not at the moment.

Uh, we're, we've-

Jethro D. Jones: with

Mike Lawrence: vibe coding.

No, no, But we've brought in some speakers that are, and they've pointed us in that direction.

Uh, and we feel like the first step for us as a district is with vibe coding.

Um, but there are some interesting barriers.

We, uh, every public school district in the country, every public agency in the country has until April of 2027 to make sure that every website, app, video, PDF is read- screen readable and accessible for all parents, family, community members.

This is an important thing.

The problem is vibe coding and some of these other tools that we would roll out don't necessarily make the most accessible Mm-hmm … and, and, and publications and videos.

Yeah.

So that's become a stumbling block for that step, and I saw that as a step towards, uh, agents, uh, and, and leveraging those tools.

Um, there is also a little bit of a fear in terms of what happens when we let those loose, uh, in terms of student data and student data privacy.

Uh, that's a deep concern of

Jethro D. Jones: for

Mike Lawrence: public school district.

Uh, and so we're doing the research on it, but we haven't really jumped into it.

I will say we do have in our district at Whitney High School, we have an AI pathway.

It's been there for seven years, I think.

I think they launched in '18, '19.

And it's one of our, um, our career tech ed, uh, pathways, and it's one of the first and only in the, in, you know, in California.

Um, and that teacher, I reached out to him, and he does a lot of work in agents, and his students are doing some work in agents.

So I, I will, I will step back my answer in that there are some pockets- where it's happening, and, uh, I just wanna make sure that
we're protecting the students' data- and that we're making sure that the work that's published and created is, is accessible for all.

Yeah.

Jethro D. Jones: Yeah.

Well, it's, it's really interesting.

I do think that, uh, vibe coding is a good step towards an agent.

Because it is, uh, it is the next step on the path before that.

And once you see that you tell something, "Go make an app that looks like this and does this kind of stuff"- Yeah … then you start to extrapolate and say, "What else can it make?

What else can it do?" And that's where, um, it becomes really

Mike Lawrence: yeah, and I will step into it.

I just, uh, I-- and I'm, I'm rarely the, the, you know, the last on the bus.

I'm pretty,

Jethro D. Jones: pretty early

Mike Lawrence: But in this case, I just, I don't love the idea of someone, some agent, agent out there that could be misinterpreted as me until I have full control over what its actions are.

So that's where my caution Yeah.

And I've just got plenty on my plate at this

Jethro D. Jones: moment.

Yeah.

I hear you.

Yeah.

So Mike, how do people connect with you and learn more from what you're doing?

Mike Lawrence: think LinkedIn is a great place.

I, I spend time on there.

Uh, I, I have a, a profile page up at mikelaurence.me.

Uh, ABCUSD is our school district website, and it's, uh, it w- if you just add a /AI, you'll find all of the initiatives that I've talked about and more.

Uh, video, um, so it's abcusd.us/AI.

Jethro D. Jones: Excellent.

I

Mike Lawrence: have stickers.

If anyone's here at ISTE, I can Yeah, great … a with that logo on

Jethro D. Jones: Good.

And I bet if they sent you a little email, you might even- find a way to get that sticker to them.

Um, all right.

Well, Mike, this was awesome.

Thank you so much.

You bet.

and tell us about your Ted Lasso podcast.

Mike Lawrence: Oh, this starts, uh, July 1st.

So if any Ted Lasso fans out there, uh, there were three seasons, and they said there were only gonna be three, but then they changed their mind and season.

Yeah.

So, uh, my colleague and I, um, had done a session on, uh, Ted Lasso that was f- somewhat popular.

The, the rooms would get closed up, room

full.

And, uh, we decided to retire that after the third season was over 'cause we were doing a different, uh, presentation every season.

And so when we finished that third season, we, we shut that down.

And now that there's a fourth season, we thought, "Well, gosh, we should do a fourth version of our session." But we don't know what's in the season yet- hasn't been launched until August 5th it, it, it premieres.

So we thought, "Well, what could we do to generate excitement and start thinking about Ted Lasso again?"

Well, we have all of these, you know, facts and, and research that we've done and how he's modeling, uh, leadership and empathy and all that.

Let's put it into a podcast.

And so we're gonna do a rewatch where we watch an episode a day, and then we get up… 'Cause we don't live near each other, so we're
gonna set it up via Adobe, um, Podcast we can connect with each other virtually and record 15 minutes a day on that particular episode.

So we've already done a teaser and we just, uh, rewatched Season 1, Episode 1.

we're gonna record that today or tomorrow, and then we'll publish that on July 1st.

And try and get a backlog going, and then we'll just publish them timely.

You know, we're trying to think like five or six every day P- Pacific Time.

And then, uh, you can watch the show and then listen to the 15 minutes of

We'll have links to social-emotional learning, leadership examples, obviously education examples, uh, but we wanted it to be a broadly interesting topic for anyone who's a Ted Lasso

Uh, there will be some spoilers 'cause, I mean, you've had five, six people.

Yeah.

Um, so if it's your first time watching through the series, uh, we're gonna try to keep those light spoilers.

Um, uh, but you might wanna watch all three seasons and then come back and listen if you re- truly don't wanna be spoiled.

good.

but, uh, but yeah, that's, that's the podcast.

It's at, uh, tedpotential.us.

Jethro D. Jones: Tedpotential.us.

Excellent.

Okay.

Well, Mike, this was awesome, and I hope people listen to that podcast.

Yeah.

I know I will.

Okay, Cause I'm a Huge

Mike Lawrence: I'll give you a bookmark 'cause there's books in every episode Beard, Coach Beard is reading.

So we put, put together a

Jethro D. Jones: Oh, check this out.

Mike Lawrence: The podcast- Very good … is called Smells Like

Jethro D. Jones: Smells Like Potential.

Very good.

All right.

Thank you very much, sir.