Episode 15: Code.org

[#1]

Ryan Lufkin (00:08.113)
Hey there and welcome to Educast 3000. I'm your co-host Ryan Lufkin.

Melissa Loble (00:12.352)
And I'm your co-host, Melissa Loble. As you know, Ryan and I have been deep into AI over this last year, and equally deep into innovative ways to really support teaching and learning happening every day in the classroom. Well, we've got a special guest today who's going to help us dig into both of these areas. We have had the privilege of meeting our guest.

through some really great work that we've been doing with Code.org. So Tess O'Brien, Senior Product Manager with Code.org, welcome to our podcast.

Tess (00:46.138)
Thank you so much for inviting me. I'm thrilled to be here.

Ryan Lufkin (00:49.086)
It's awesome to have you. Well, and Tess, before we dive into questions, give us a little bit about your background.

Tess (00:55.12)
Definitely. So I'm a senior product manager at Code.org. My career up until this point is focused on how we can bring technology to nonprofits. I work with a small group of engineers and educators at Code.org focused on our teacher initiative. And I was one of the first folks who was reassigned about two years ago to work on generative AI products. I spent the last two years building our AI teaching assistant, which was just recently.

awarded the Best in STEM AI tool by a group of educators.

Ryan Lufkin (01:28.308)
Which was actually, we saw it it's one of reasons we wanted you to be the shows.

Melissa Loble (01:32.44)
Yeah. Tess, how did you find co.org? I'm just kind of curious. What brought you to co.org to begin with?

Tess (01:39.27)
So a little bit of background about me. I was raised by like matriarchs of my family. My mother and my grandmother came over from Cuba and my grandmother was actually a really talented mathematician and was one of the first women to work at IBM when she came back to the States and so really neat. But I sort of was, I was raised with

Ryan Lufkin (02:00.607)
Yeah, that's cool.

Tess (02:06.884)
this notion of this woman in computer science who I believe was a big part of our family's security and financial stability and seeing that role model was very foundational for me. And then when I had the opportunity to work at Code.org and to help change the face of who was in computer science and who feels comfortable in a computer science classroom, felt like a really great way to honor her. So I've been there for four years now and it frankly is my dream job.

Melissa Loble (02:36.271)
That just gives me chills for so many reasons. I love that story. And we're going to dig in a little bit more to what Code.org does. But I love the connection. mean, Ryan and I are very passionate about this, too, of helping to have impact on learners every day, everywhere. I wonder if, so the other question that we typically ask our guests is, what's a favorite learning moment? And so I wonder if I'm kind of guessing it might involve.

Ryan Lufkin (02:49.738)
Mm-hmm. Yeah.

Melissa Loble (03:05.285)
someone in your family? Maybe not though. Do you have, would you mind sharing like a moment either where you were a learner, you were a teacher, you observed learning or a favorite learning moment?

Tess (03:15.558)
Absolutely. I was reflecting a little bit on how to pull together threads that I think we're talking about here, both with Code.org and with this new space of generative AI and education. And what came to mind was when I was first introduced to the concept of growth mindset as opposed to fixed mindset. So despite having worked in the tech industry for almost a decade now, I did not take a computer science class until after college, so never was exposed.

Ryan Lufkin (03:43.744)
Wow, yeah.

Tess (03:45.602)
I worked for nonprofits and as I was getting into volunteer management work, I realized just how much of the field was really rooted in technology and felt discouraged that there were career doors that felt closed to me. So I went to a software development immersive course where you go from never having coded before to in theory being industry ready after three months.

Melissa Loble (04:00.514)
Mm-hmm.

Tess (04:12.9)
And I was sitting in that first day, like setting up my coding environment and in the back of my brain, like screaming to get out of this room, trying to like scheme how like, can I get the job that I just left back? Like, how can I save face? Like I don't belong here. I've had these like deeply ingrained, totally untrue notions. Like I'm terrible at math. What am I doing in this room? Like this was a rash decision.

Ryan Lufkin (04:20.758)
Ha ha ha ha ha ha.

Melissa Loble (04:21.174)
Mm-hmm.

Ryan Lufkin (04:36.322)
You

Tess (04:42.626)
And before we left the classroom that day, the instructor gave everybody a reading to do on growth mindset versus fixed mindset. think with the notion that there might be a couple of people in their own feelings. So we, this reading laid out the idea that all learning is cumulative and you learn by failure and you learn by effort and resilience. And while it sounds inherent, it just like isn't something I had applied.

Melissa Loble (05:05.946)
Mm-hmm.

Tess (05:12.144)
to myself before. And as opposed to this idea of fixed and so like I'm good at math, I'm creative, like I'm naturally something and I certainly had fallen into that trap. So from that moment, it was kind of like cut and run and like swallow my embarrassment or what if I just radically believed in a growth mindset for three months for this period and just made it like my core belief.

Melissa Loble (05:13.871)
Mm-hmm

Ryan Lufkin (05:21.176)
Yeah.

Melissa Loble (05:31.906)
Mm-hmm.

Tess (05:42.658)
And I went that path and it turned out really great. I've like obviously gotten a job and built a career in this space, but even more so it's something that has been ingrained in me and the work that I want to do and the work that I get to do at code.org, both with students and teachers.

Ryan Lufkin (05:59.715)
I love that focus on that growth mindset. know, the imposter syndrome really is real. I think especially when we're trying to learn new things, it's so easy to get discouraged at the beginning and think, man, I'm not good at this, right? And you don't know it yet. You've got to learn it, right? And that's, it's all about that mindset to overcome that. Amazing.

Melissa Loble (06:14.266)
Mm-hmm.

Melissa Loble (06:17.862)
Mm-hmm. I love it too. And we're going to talk in a little bit about pulling that into the work that Code.org does as well. I see so much of your connection with Code.org and then also your leadership and vision. But not everybody listening to our podcast is going to know who Code.org is. So could you give us a little bit of background on, introduce us to its work?

what it's doing now, and sort of the impact Code.org has had on education more broadly.

Tess (06:48.228)
Yeah, absolutely. So co.org is a nonprofit that was founded on the idea that every student in every classroom deserves a chance to learn computer science. So I work in the product arm of the organization, which focuses on bringing free curriculum, professional learning and tools to classrooms around the world. Outside of that, the organization spans much farther. So everything from

an international arm that is bringing computer science curriculum to the world, to an advocacy and policy arm that is focused on passing computer science requirements to help get more students access to this education, to movement work to help spread the word about the computer science movement. In terms of reach, we have reached 92 million students worldwide, with 2.7 million teachers in 180 different countries.

Ryan Lufkin (07:40.006)
That's amazing.

Melissa Loble (07:42.394)
Those are big numbers. So we just have to say it. So how many million students? 92 million. And how many teachers?

Ryan Lufkin (07:44.241)
Yeah.

Tess (07:48.006)
92 million. It's pretty shocking.

2.7.

Melissa Loble (07:54.566)
Crazy. I just wanted you to repeat those because that's a big impact of an organization focused on computer science, but it's having such a bigger, reach than that.

Tess (07:56.71)
Yeah.

Ryan Lufkin (08:06.888)
Yeah. And actually I have to admit, years ago, I'm trying to think how many years ago it was, probably almost 10 years ago, we actually used some of the Hour of Code resources for a bring your kids to work activity we did at another company I worked for. And so it was amazing because they were bite-size activities that students could really dive into. really it made coding accessible.

Melissa Loble (08:20.922)
Mm-hmm.

Ryan Lufkin (08:35.59)
Right, it made it seem like, my gosh, I'm doing these simple tasks, but look at these results I'm getting, right? And my daughter at the time, you know, she's now 20 and 10 years ago, she, I was hoping for a STEM child. I may still end up with a STEM child. But it was really interesting for her, because it made it a possibility for her, something she hadn't considered previously. But what is it about code.org? know, Melissa and I, Melissa mentioned a little bit that we very much, there's a,

an aspect of reward that comes with working and education that you feel like you're having an impact on the world. What is it that gets you excited every day about Cordelari?

Tess (09:13.798)
Great question. think it comes down to, I'm at everything, but it comes down to. So specifically I run the teacher initiative at code.org and computer science teachers are by no means a monolith. So we have everybody from industry professionals who found a second career in education, to folks who were voluntold a few days before school started that they are now a computer scientist.

Ryan Lufkin (09:19.729)
Hahaha

Melissa Loble (09:19.961)
Mm-hmm.

Ryan Lufkin (09:39.465)
Yes, I've had friends in that that voluntold space, yes.

Tess (09:46.054)
And it takes a, for those folks, but for all computer science teachers, it takes a pretty tremendous amount of courage. notoriously, computer science is constantly changing and as an engineer, you work your eight hour day and then you work two more hours to learn the thing that's going to be next. Computer science teachers have to do that same thing. And so my team gets to be hyper focused on computer science teachers and how can we save them just a little bit of time? How can we make

their career just a little bit easier. How can we bring them confidence and ease in their work? Which is, I don't know, a really good reason to get out of bed. And the other piece is that scale. Being a free curriculum operating at that scale, there's this really high bar for quality that you want to bring to your work and innovation that you want to bring to your work, especially because some of the students we reach, because we are free, we are the only way they would get this curriculum.

Ryan Lufkin (10:25.318)
yeah.

Melissa Loble (10:27.695)
Hmm.

Ryan Lufkin (10:36.787)
Yeah.

Tess (10:45.264)
So it makes it like there's a ton of pride in the work that code.org does.

Melissa Loble (10:51.065)
I do too. Well, and you shared when you were introducing yourself that you've been working on Code.org's AI initiatives for the last two years, particularly the teacher assistant, which we'll get to in a few questions. But before we get into how you're leveraging AI, we have all sorts of different listeners. We'll get educators and teachers, and we'll get other vendors out there in this space, all sorts of people, leaders in education.

And it's always interesting to unpack some fundamental product principles, especially since you are a leader in your product organization, around how and where you use AI. So there's a lot of debate around AI. Is it harmful, more harmful than it's good? What are some of your foundational product principles around AI, or what are code.orgs, or both?

Tess (11:42.03)
Amazing. So I think to start, I'll go broad and overarching philosophy I try to hold and then hyper-tactically I can get into some like real bullet points of things we've learned. So broadly, I think I like many things in life. I've learned that AI product management or a product development is really holding

Ryan Lufkin (11:53.29)
Well, we could we don't go down that rabbit hole for sure, but yeah

Melissa Loble (11:57.712)
I love it.

Tess (12:10.726)
contradicting ideas at the same moment. So it's this abundance mindset along with this caution mindset. To dig in a little bit more there, the problems that felt unsolvable before or felt like they would be too expensive to solve, like, well, I'll never have enough engineers to do this. All of the sudden, there is this new amazing tool that might be really helpful and might allow you to dream that big. The flip side is there are

fewer industry standards, like less best practices because it is such a new technology. And so to some degree you are developing in the wild west and you have to put your own cautionary boundaries and move with extreme caution knowing that it is new technology and there's so much that is unknown about it. So I think it's always just kind of trying to dance between those two to get hyper tactical with things we've learned and.

policies that code.org has put in place around this. I would say five of them. Find the right size problem. So like all things product management, to find a really specific problem that is painful and AI is the right tool to use and solve. Next, I would say humans in the loop. This is a classic AI benchmark to be using at all times.

Ryan Lufkin (13:22.88)
Yeah.

Ryan Lufkin (13:32.276)
Mm-hmm.

Tess (13:34.502)
But for example, thinking about assessment, you would never want to send an AI score to a grade book or you would never want to send an AI score directly to a student. But sending that score to a teacher to help them assess and help them save a little bit of time in the process is something that we felt comfortable doing. So making sure that humans are verifying the AIs. Next benchmarks. So setting your own benchmarks for

Ryan Lufkin (13:39.296)
Yes.

Ryan Lufkin (13:55.945)
Absolutely.

Tess (14:02.576)
for each product. Code.org, use quality, or we use accuracy, safety, and bias. And for each product that we launch, we think about each of those. We set a bar we're trying to hit, and we work until we hit it. Next, I would say plant a garden. So there are so many ideas of things you could be doing. And ultimately, the technology is going to get there. And so fail fast.

Ryan Lufkin (14:16.758)
Awesome.

Tess (14:29.84)
Create proofs of concept. If the technology isn't there to hit your benchmarks yet, hold it gently elsewhere and know that you can come back to it as this tech gets better. And finally, I would say look for help. So while there are lots of small teams around the industry working on AI products, which could be really lonely, also those teams are all going through the same problems at the same time and coming up with their own unique solutions and knowing

the computer science community, knowing the education community, they really use this spirit that I've never seen before of sharing and collaboration and fellowship.

Ryan Lufkin (15:11.439)
Those align so well to our principles, what we're telling our community. And that's one thing we've seen in the instruction community is people coming together to learn from each other. You're not starting from scratch. If you're trying something with AI, you're trying something new, somebody's already done something similar and wants to share, right? So I love that you're seeing the same thing. It's incredibly exciting that, especially in education, educators support each other so prolifically.

So I love the commitment to AI. How does that translate into how you're actually applying AI into your efforts?

Tess (15:48.256)
Absolutely. So AI is part or next to the computer science field in general. So the moment GPT 3.5 came out, we realized this is going to change the education industry and we need to support our users through this. Code.org's overall approach was three pronged. So the first was coalition building. We launched Teach AI, which is an organization that brings together tech education leaders with

Ryan Lufkin (15:55.682)
Mm-hmm.

Tess (16:19.078)
classroom leaders and district leaders and helps to guide to build safe and effective use of AI in classrooms. Next, we realized that because this tech was so new, there was a big vacuum in both curriculum and professional learning. So code.org and other folks had more classic AI training available, but this generative AI curriculum, we needed to move quickly to right curriculum and professional learning to support.

And finally, and this is where I sit, we started applying using this technology in actual tools for our users. So my team was building the AI teaching assistant or has been, and I can dive in a little bit there. It's my favorite thing to talk about. So we, to that abundance mindset and solving big problems, we thought about what things that we never really have the resources to solve in the way we wanted to.

Melissa Loble (17:01.17)
Please tell us more.

Ryan Lufkin (17:01.636)
Yeah, definitely. Yeah.

Hahaha.

Tess (17:16.746)
And the answer is very clear. Our teachers every year answer the question, what are the biggest pain points for you? And differentiation, being able to adjust curriculum for individual students and assessment, knowing where a student is and understands what they don't understand are constantly the biggest pain points, both in teaching computer science and in using code.org. So we thought, what if we could build a tool that helps with these

pain points. So we started with assessment and found a really specific problem. this problem... I'm gonna talk again, sorry. So we started with a really specific problem. Computer science is project-based, meaning that students work on projects that they can solve in a billion different ways.

Ryan Lufkin (17:58.866)
We can edit that out.

Melissa Loble (17:59.218)
Yeah, go for it.

Tess (18:13.626)
And what makes it really difficult to assess is that teachers need to read each student's project. Each student's project will be different and figure out what they know and what they don't know. Generative AI is actually really great at looking through code and understanding what it's doing and understanding where it might fall short. So we built a tool that sends student code and a rubric to an LLM and returns scores for teachers.

Teachers before this product were reporting up to 50 hours of grading per unit per classroom with 10 to 15 minutes per project. After we launched this product they're reporting up to 50 percent time saved.

Ryan Lufkin (18:47.748)
jeez. Wow.

Ryan Lufkin (18:59.033)
Wow. That's cool.

Melissa Loble (19:01.02)
That's incredible. And like you said, a really good use of AI, a specific problem, and something that it's good at. It's mapping content or mapping code, right, and rubrics against outcomes that then the human can lean in. I've seen this, by the way. It's super cool. And then decide where and how they want to leverage the feedback that's coming back from AI. By the way, for our listeners, we will

Ryan Lufkin (19:17.115)
Mm-hmm.

Ryan Lufkin (19:23.569)
Yeah.

Melissa Loble (19:27.22)
Make sure to include in the show notes links to Teach.ai so you can see those resources, as well as links to be able to see this really great teaching assistant that Tess is talking about. So this is a great foundation. What's the future? Where are you looking at co.org and how you either are using AI or just more broadly around the curriculum you create, around the technology tools that you create? What's the future look like?

Tess (19:53.09)
Absolutely. So on the curriculum side, we have a team that is constantly building new curriculum based on generative AI. So we're building curriculum to help users and students know how to use AI as well as how to build AI. So supporting all students and incorporating AI into their workflows, as well as supporting students who maybe are interested in learning how to build these things. Additionally, on the how we are using AI to support teachers.

Ryan Lufkin (20:04.835)
Mm-hmm. Love it.

Tess (20:21.838)
I mentioned assessment and talked a bit about assessment. We've moved on to differentiation. So once we know what a student knows, how can we adjust the curriculum to support that specific student? We believe if we can bring these two things together, we can really change the learning gaps in computer science. And so I think the future of the AI teaching assistant to go back to that abundance mindset is we can get this right. If we can measure what a student knows.

Ryan Lufkin (20:39.41)
yeah.

Tess (20:51.29)
We can then adjust curriculum to meet that student's specific needs. So future big dream, where we're headed is curriculum that adjusts for a specific student, both what they're interested in, who they are, and the things, the misconceptions they may have to really help all students thrive in computer science.

Ryan Lufkin (20:56.155)
Yeah.

Ryan Lufkin (21:12.303)
That's amazing. And we've all had that experience where maybe we just didn't connect with the instructor or, you know, and because they did it a little differently than, you know, made sense to us. And I've seen it with my own kids now. So that's amazing being able to adjust that and provide a personalized experience. When we met a couple of weeks ago, you actually shared some of the research you all have done around connecting computer science skills with broader student success. Do you want to share a little bit about that and kind of, you know,

Melissa Loble (21:17.18)
Yeah.

Ryan Lufkin (21:42.097)
even for non-computer science students, how it's beneficial.

Tess (21:46.608)
Yeah, absolutely. So students should take computer science because of the computational thinking skills and the problem solving skills that it affords them. Not every student who goes through a computer science class is going to be a developer or work in tech, but we believe that those skills are foundational enough and computer science is one of the best ways to learn them. Recently, a study came out that really backs this up. So Brown University did a study linking the effects of computer science courses in high school.

with tracking the same group of students through adulthood. And they found that there is an increase of wages of 8 % overall for students who took a high school computer science class. It's pretty big for us.

Ryan Lufkin (22:28.401)
Wow. yeah, that's pretty substantial. And we'll link to this study too, by the way. We'll include this in the show notes as well.

Melissa Loble (22:35.6)
Yes.

Tess (22:37.848)
Absolutely. sort of the kicker here is that those gains are even higher for underrepresented groups. So young women see even higher gains. Students from lower socioeconomic status groups see lower, to see higher gains, and black students see higher gains as well. So we have this super encouraging data that tells us that this works.

But we are also at the same time seeing that only 3 % of young women are taking computer science courses, despite 82 % of students having access. So we are really hyper focused at code.org right now and meeting these students where they are and helping them understand that computer science is for them. There is space in the classroom for them and that we need them in this industry.

Ryan Lufkin (23:26.831)
That's fascinating, that.

Melissa Loble (23:28.208)
I love it too. It speaks to me. I know I shared this story with you and Ryan, but I'll share this with our listeners. So in seventh grade, I went to public middle school. And then seventh grade was the first year that they offered a computer science course. You could take that or typing. So I have subsequently never learned to type because I took that computer science course. And I was the only girl in the class that opted into that elective. They didn't offer after that, so I couldn't continue to take those.

Ryan Lufkin (23:47.513)
You

Melissa Loble (23:57.8)
But to this day, I remember that class. And it makes me wonder, also being a practitioner that's not an engineer by trade, but that has been in the technology space for a long time, it makes me wonder, again, those skills that you described, how much of that came from even just thinking differently from learning basic DOS, which was super cool. And I remember it was the first time I'd ever experienced as a young girl.

Ryan Lufkin (24:19.749)
Yeah.

Melissa Loble (24:27.836)
being the only girl in a setting like that, in a learning setting, which all your other classes, you're going to have at least one other girl. And typically, you're fairly balanced. If not totally balanced, you might be a little underweighted on the girls, at least when I grew up. But it's wild to kind of think about that and how that framed sort of my world. What I love about CodeDog.org and the work that you do is

Ryan Lufkin (24:31.291)
Yeah.

Melissa Loble (24:53.232)
you think about things like this. What is the bigger societal impact that we could be having? How do we lift up everyone? So you've shared a little bit about the future of the teaching assistant, the future of how you're using AI. Anything else that you want to share that Code.org might be cooking up for the future?

Tess (25:12.258)
Yeah, absolutely. And first, thank you for sharing that. Some of the best parts of this work is getting to find shared threads amongst other people who have found.

Ryan Lufkin (25:21.521)
Mm-hmm.

Tess (25:24.522)
some part of their identity or confidence in these classrooms and do problem solving. Asking me that question this time of year, Ryan, already started talking a little bit about this, but it is Hour of Codes, using netcode.org. So this is our 11th Hour of Code, which is founded on the idea that if students just spend one hour coding, they'll get that foot in the door. They'll have their first chance to experience computer science.

Ryan Lufkin (25:27.6)
Yeah.

Ryan Lufkin (25:48.42)
Yeah.

Tess (25:51.768)
In the past, we've had everyone from presidents to popes to my nephew Charlie's first grade classroom do an hour of code before. And this year we have a lot planned for it. So we have new tutorials, I think I can say focused on AI for folks to use. We have great partnerships, including one we just launched with Shakira as well as professional learning opportunities to support teachers who maybe

Melissa Loble (26:07.804)
Mm-hmm.

Ryan Lufkin (26:13.553)
Ooh, that's

Melissa Loble (26:14.247)
Cool.

Tess (26:21.51)
want to see what this is like. So I would say overall, educators, administrators, folks in higher education, parents, friends, all things, Hour of Code is a great moment to sort of dip your toe in that pool. So I would say right now, top of mind for us is check out code.org, host an event, host reverse Hour of Code and get a sense of what this movement is all about.

Ryan Lufkin (26:49.873)
Yeah, and I actually will link to that because I honestly, the hour of code when I discovered it, I was like, well, this is a treasure trove. That was even before AI, right? I mean, so it was just introducing students to the basics and getting them interested in really fun ways was, was super compelling. But as we talk about AI, AI literacy is something that Melissa and I spent a lot of time talking about. Obviously, code.org and you yourself are pretty passionate about AI literacy. What can we do together collectively to really

improve AI literacy, I guess not just for students, in the broader society, right? It's something that is increasingly going to affect everyday life in pretty significant ways.

Tess (27:31.716)
I love this question and reflecting on it, like frankly, I the work you're doing both with your larger community and with opportunities like this to just allow people to have open conversations about AI that start from the space of.

No one knows everything about it right now.

Ryan Lufkin (27:49.039)
Yeah, yeah. And it's moving fast, but you're not too far behind, right? That's the other piece too, is I think sometimes people feel like the train's left the station and they're still on the platform. And I encourage them to be like, no, no, no, it is not, it is not.

Melissa Loble (27:53.916)
Mm-hmm.

Tess (28:02.054)
And not one bit. the ramp up is like so quick with generative AI. Like it is so possible to quickly ramp up to like hyper tactically again, to answer your question. I think we need coalitions like TJI and other folks that are bringing people together in this space, like Instructure, to help push for common sense policies that are a little bit more unified.

Ryan Lufkin (28:07.728)
Yeah.

Tess (28:30.118)
When I go to conferences and talk to teachers, especially last year, the loudest thing I heard is we need district guidance on what we should and shouldn't be doing. And so helping support that district guidance. I'd say the next piece is really up-to-date professional learning and easy, lightweight professional learning. Because this technology is moving quickly. so keeping it up to date and learning about things as they come out.

Melissa Loble (28:49.372)
Hmm

Ryan Lufkin (28:51.472)
Yeah.

Tess (29:00.414)
And then I would say finally having vulnerable conversations at all levels about this technology, its pitfalls, its strengths. I was speaking to a teacher, Renee, who uses our AI teaching assistant recently, and she took the time that she saved by using AI to assess her students' code to sit with each student in her classroom and show them this totally teacher-facing tool and talk.

Ryan Lufkin (29:09.776)
Yeah.

Tess (29:28.742)
through them each answer. And when it was wrong, talk about why maybe AI was getting it wrong and pick that apart. And then to get us feedback on each of these scores. And I think that's like a microcosm of the type of conversations we all need to be having.

Ryan Lufkin (29:33.551)
Wow. Yeah.

Ryan Lufkin (29:44.996)
Yeah.

Melissa Loble (29:46.28)
That gave me chills. Let's talk about, Ryan and I will often talk about how AI affords us the opportunity to be more connected with our students, not less. And I think so many people are afraid it's pulling that human out. It's actually putting the human that much more in, in your example, which is just like, love that for so many reasons. I have one last question, and I know Ryan may or may not as well.

Ryan Lufkin (29:47.473)
That's incredible.

Ryan Lufkin (30:06.321)
Yeah.

Melissa Loble (30:14.78)
Where do you predict, thinking about that, in general, where do you think AI will be in the next three to five years in education?

Tess (30:24.086)
so wild to think about, like certainly a market. Reflecting on it, I feel like I can say with a somewhat high degree of confidence, surprisingly, where code.org will be, which I think is sort of a harbinger of how all of these industries and districts are starting to get their arms around this. So I believe that code.org five years from now will be.

Ryan Lufkin (30:26.523)
Hahaha.

Tess (30:52.166)
closing learning gaps in computer science using this technology. I believe that will be customizing our curriculum for individual students at scale. I think that reflects on the industry and education as a whole to some degree of we've been experimenting up to this point, but we know those big broad ideas we want to solve. We know the biggest challenges that we want to face. And so I think that over the next few years, we're going to start seeing tools that

really make teachers' lives easier, really have impacts on learning. And at an individual level, Ryan, to your point about this train has not left the station, I think we're only now at the point where we actually understand how generative AI can be used for productivity gains. There's been about a year of research on it, and we're gonna start building better tools that don't involve you to becoming an expert in generative AI to get the benefits of it.

And so think teachers are really gonna start to see those gains.

Ryan Lufkin (31:51.136)
Yeah, I think it's funny because, you know, people would say, well, I want to be the one writing poems and drawing pictures, not AI, right? And I think we really have shifted in the last, I'll say in the last, you know, six months to a year, we've shifted into this, okay, now how do we really put these tools to work productively instead of doing the simple stuff, which was really interesting and has been amazing to watch. Really now, I think how do we apply these in amazing ways? So I love what Quintenoria is doing. I love your work because I think you're...

Melissa Loble (31:52.05)
cool.

Melissa Loble (32:00.375)
Mm-hmm.

Ryan Lufkin (32:20.414)
you're really making all this actionable in ways we're building trust, we're building knowledge around AI, but you guys are actually putting it to work in really incredible ways. And so I really appreciate the effort. Everything you guys have been doing, so amazing.

Tess (32:34.074)
Thanks so much. Thank you. again, thank you for creating spaces like this to have these types of conversations.

Ryan Lufkin (32:41.054)
Now Tess, you've been amazing. This is exactly the kind of conversation, you know, I feel smarter every time we talk to our guests that like this, and certainly I walk away from this feeling smarter and more excited about, you know, the future with AI and teaching and learning for students. And I don't have any more questions. Melissa, do you have more?

Melissa Loble (32:48.033)
Yeah.

Melissa Loble (32:58.425)
No, I think we've picked Tess's brain enough, but we will put lots of links in the show notes as Ryan and I both mentioned on some of the resources and references and research that that Tess mentioned and how to get involved with Code.org. I mean this is such a it's it's such a gift in education whether it's K-12 education, community college, wherever you might be leveraging this your own personal journeys. It's such a gift so we'll make sure we get to share it.

Ryan Lufkin (33:00.542)
Hahaha

Ryan Lufkin (33:16.074)
Mm-hmm.

Ryan Lufkin (33:23.872)
Amazing resource. Tess, thank you so much for joining us. You've been awesome. And I'm sure we'll have you back again in the future to, you know, catch up on what happens between now and then.

Tess (33:32.934)
Thank you so so much for having me. This has been a blast.