A Conversation with Dr. Yohan Lee
Every day, there is a new headline about AI. Companies achieving breakthroughs. Companies getting sued. AI “going rogue” and seemingly developing a nefarious “personality.” People who are blind experiencing the world in a whole new way through smart glasses, which can recognize objects and immediately audibly describe what is in front of them. While the applications of AI appear endless at the moment, it can be overwhelming to think about. Should I be worried about how my child uses AI? Should I be worried about my job? Should I trust what AI tells me?
There are endless questions and so many conversations to be had. Today, we have the privilege of having just one of those conversations with someone who is a Christian and also has deep experience in technology and AI—Dr. Yohan Lee, Associate Dean of Technology at Biola University.
In the following interview, Dr. Lee weighs in on a small sampling of questions about AI’s applications and its impact on our current moment. This conversation isn’t meant to be exhaustive, nor is it meant to be the be-all end-all for the questions discussed. We hope it’s a conversation starter that can lead to more discussion and thought processing about our ever changing modern age, and spark ideas about how Christians can interact with our current moment in order to share the gospel with all.
This interview has been edited for length and clarity.
0:00 Intro & Background
1:32 Dr. Lee’s Faith Background
2:50 What is Artificial Intelligence (Explain to me like I’m five years old)
3:24 A discussion about AI Ethics
7:14 Some Positive Applications of AI
12:36 Help! AI Took My Job
18:04 AI and Higher Education
21:47 Technology & The Church
Introductions
You’re currently the Associate Dean of Technology at Biola University, but have had a pretty extensive career across different areas in STEM. Can you please introduce yourself and the path that led you to where you are today?
Yes, thank you for having me. It’s such a privilege to be here. Yes, I’m a Bay Area kid, a Silicon Valley kid. My dad had a few patents that were directly intersecting with microchips back in the 80s, and so that’s how I ended up here.
In terms of education, I majored in neuroscience and I did a PhD in genomics, both at UCLA, then upon postdoc, I became a government scientist at the National Institutes of Health, where I did a lot of work with robotics to systemize genomic work for rare diseases in clinical trials. After NIH, I was at Booz Allen for technical consulting, so that’s big cloud, advanced analytics, and I was part of the first class of AWS Certified architects at that company. And then fast forward, I was at Google AI, where I was a principal investigator on a number of projects, then joined a startup, and then helped that company double the revenue, expand to three continents, help them with a massive fundraising round, and then I launched my own company, which I’m still running. And the university was so kind to support me in that endeavor, while I took on the role of Associate Dean.
Faith Background
Your STEM background is pretty extensive. Can you tell us a little bit about your faith background?
So my parents, my dad, was third or fourth generation Christian, depending on how you want to count. My great grandmother, my dad’s side, knew some of the original American missionaries that came to Korea. And then my mother came to faith when she was young, and then led her whole family to faith while she was still in Korea, and then she went on to get her Masters in Divinity at Princeton. We were a Presbyterian family, so we just kind of followed that role. And so in the sense that I was well churched and grew up in a Christian home, yes, but I guess the more important part to it is, you know, my mother pressed this upon me really hard in high school when she saw, you know, I’d be going to church. But then sometimes my character didn’t reflect that, and she would say, look, faith has to be real. It can’t be an occupation, it can’t be an interest. It can’t be something that you, that you you flirt with, you know faith has to be either the center point of your life or you’re really on the wrong path and and so for me, making the Lord the Lord of my life while between undergrad and grad school was probably one of the most important moments of my life.
What is AI?
Explain to me like I’m five years old. What is Artificial Intelligence?
That’s actually really important, because there’s a lot of hype, and so the way I like to explain it is, it’s teaching a computer to do things like a human, such as learning from facts, seeing a pattern, and then guessing what could or should be next, and then doing something automatically for you that normally takes a lot of human thinking behind it, so that’s kind of my working definition.
AI Ethics
As Christians, should we be conscious of our AI usage because of its impact on the environment? What are some impacts of AI that we don’t necessarily see in our day-to-day usage, but are important to consider in terms of ethics?
Data centers are a massive capital commitment. They’re a massive investment commitment. But at the end of the day, yes, should we be considerate about how much and how often we use AI? I think so, because I am conscious of really, you know, how much more electricity and power a query on ChatGPT is, for example, as opposed to a Google search query. You know, Google, despite its faults, I mean, for 20 plus years, they’ve done everything they can to be more sustainable, more capable and more efficient. In fact, one of the earliest algorithms for AI that they used was to reduce the energy utilization and increase efficiency in their data centers. I mean, they did that, not because they had to, because they chose to. And so that’s an example of being extremely responsible with what AI is and what it takes.
I guess the other question is, it’s really and this is my invitation to folks who are Gen Z, and that’s a lot of them rely solely on ChatGPT as their first touch for the internet. Like you said, you know, you can ask a query on your phone. You can re-ask that query right away. Sometimes it’s on their augmented reality sunglasses or eyeglasses where they immediately interact directly with ChatGPT as that first line of electronic Web3 touch. But the part with that is, while I recognize that that provides tremendous convenience, the greater question that you’re getting at is, is that the best usage of resources that we don’t see and touch, especially in a developed economy, but for other folks, it has a direct impact, and so I think more of that does have to be explicated.
I also like one of the items that Karen [Hao] highlighted in her book [Empire of AI] about folks outside the US in developing nations that are essentially used as a data labeling economy. In some regards, in some places, that is a lifeblood to an economy that hasn’t had opportunities, it’s wonderful, but in some areas, it’s unfortunately exploitative, because in content moderation, that’s, you know, reviewing content that is objectionable. At the end of the day, it’s a human that often has to identify and label that despite AI recognition tools and that causes tremendous psychological and emotional trauma, where some companies that are responsible have actually had to create wellness and therapy to support their employees who get exposed to too much of this material.
One of our professors here, Professor Ed Byun, who did the [Save My Seoul] documentary on young ladies in South Korea under sex trafficking. You know, he was reminding us that there are some people, when they’re exposed to horrible graphic material, can actually absorb it, process it, and then leave it outside, essentially, kind of like it’s part of the job. You do it as part of the job, but then you leave it outside. You don’t bring it home with you. There are some people that are naturally capable of having that type of psychological resilience, but then there are people who do not have that particular type of psychological resilience. They may be stronger in some other capacity, but when they come home, it haunts them. It plagues them. It destroys a piece of their their heart and and so the way that Professor Byun talks about it is if we were truly being ethical and moral, meaning, you know, the right and wrong, but now, as a moral person, you’re doing the right thing, we should be screening these candidates for their affinities. Are they the kind of person that can process this and leave it outside, or are they the person that’s going to process this and become inordinately more affected and impacted by it. That’s the kind of approach that I would take, so that we’re doing right by humans all over the world, and being completely irrefutable in being both ethical and moral.
AI’s Positive Applications
Even though there are those who use AI in ways that don’t build up others and can be harmful, there are many positive applications as well. What have been some of the best applications of AI that you’ve seen which have empowered people?
Where AI, I think, is tremendously wonderful, and where it should be used heavily by experts is one is in telecommunications, right? So telecommunications, which powers the world right now, is resource intensive, but one thing that I saw was an AI that helped engineers develop a form of a 6g Wi-Fi antenna, where it used the methodology and a patterning that, according to historic methodologies, was completely outside the box. Essentially, this is where machine learning can be incredibly effective, because it can identify a route to an outcome, desired state, without the historicity and the intuition and the human biases that we may have had on how certain things need to be built. And they the engineers admitted that what would have taken normally weeks to design the AI was able to do in a fraction of the time and came up with circuit design patterns that no one would have conceived of, but then, upon heavy testing,
look like the initial evidence looks like they could be more efficient than what they had initially desired.
So that’s a great example, right? And it’s not because I’m trying to create more opportunities for more interconnected objects and more addiction to phones. No. The reason why I care about things like 6g advancement is because then again, to parts of the world where they are under resourced, then the trickle down effects from 6g in the developed economy then will lead to more cost effective availability of like 2g, 3g, 4g in developing nations. And so innovation actually produces dividends for folks that are in underdeveloped nations. That’s just a reality. It’s an economic reality that’s been proven multiple times. And so innovation does matter, especially in an economy that’s going to demand faster, better, quicker, then it can help those that don’t have access.
Another one that I care about a lot is machine translation. So that’s the ability of using AI to translate text from one tongue to another. And that’s important because, you know, Meta admits that when they made one of their original models with hugging face, they trained it on the Bible, because the Bible is the most heavily translated document, dense corpora known to man, and that’s important. And so when I speak with missionaries and pastors and spiritual development leaders, if you will, and when they are trying to expand the gospel in different native tongues in different areas. You know, the Museum of the Bible tracks this very, very well, how many translations of the full New Testament, the whole Old and New Testament exist for native tongue. There’s still a huge deficit.
I believe in Sola Scriptura. You know, the power of the Bible is, in and of itself, innately and sufficiently amazing and capable of doing things. But when, when, when pastors in other countries who don’t have the resources of commentaries, devotionals, explanatory text in modern colloquial instance, that’s really challenging for pastors. And so I have met pastors from different parts of the world who say we need more supplemental material in our language to help people in our areas where they may not have had a formal education, but then to relate the meaning behind some of the parables and passages in Scripture, which, out of context can be bewildering. And so that’s an area where machine translation in a colloquial context, in an informal context, in a cultural context, can be incredibly helpful.
My mother, she’ll send me Bible verses and a context piece on how I can use it for folks in different places, and even with the best machine translation tools, and South Korea is one of the most internet informed and internet resourced nations in the world. When she sends me these things, and when we see the translation, we laugh, because it still gets certain very, very important, nuanced things completely wrong. And if you’re trying to serve somebody who just went through trauma, or serving somebody who is going through a tremendous life upstaging, you can’t get that wrong. You need to be able to convey that accurately and correctly and so that’s another area.
And then the last one is healthcare. Same thing, you know, part of the reason why Korea, you know, was able to come out of post war trauma so effectively was American missionaries who came and helped build hospitals right away. You know, you need to serve the practical needs of folks right away, to not only engender trust, but to show that you’re not completely cerebral and completely spiritual. You know, we are given these God-honoring bodies, and there is a theology of the body, right? And so that means serving one another in terms of their infirmity, their illnesses, their disease. You know, we like to here at Biola, we like to serve from a missional standpoint, that means providing clean water in parts of the world where clean water is high, hard to access, and very expensive, and so we see that concrete being hands and feet a means of serving in ministry.
Help! AI Took My Job
In Genesis, work is cursed as a result of the fall. In a sense, AI is reversing the curse by making work easier, and yet there have been mass layoffs. We were created to work, but because of AI’s impact, many people ironically cannot find work right now. One could also argue that this is also contributing to a deeper “class” divide. Do you have practical advice as to how to navigate this for those who might find themselves in this situation?
Yeah, I mean, from, I’m from the Bay Area. I’m a Silicon Valley kid, and ironically, the jobs that are being most affected right now, the biggest layoffs are coming from tech, where, essentially we are seeing layoffs in the order of 250, up to 1000s, from a lot of the big tech players. This year, as much as 10% of an entire workforce is being replaced with an AI that can support in customer service, for example, at Salesforce, etc., so that that gets real very quickly, right? So both tech and knowledge economy jobs are being drastically affected. So, is it creating a class divide? I think the answer is yes, because on the one hand, practically students who graduate who have AI skills, who know how to use AI effectively and efficiently to be able to accelerate their work output, their deliverables faster, you know, quote, unquote, more polished way. Yes, they are going to get the jobs, as opposed to students who do not have those things.
But then, let’s say you’re on the other side of the equation. It’s like, you know, you’ve been in tech for 20 some odd years, you’re a seasoned professional, and you just got laid off. What do you do? Right? And I know senior managers who have experienced this and are experiencing this, they ask me, okay, Yohan, what are the things that I need to read up on? What are the things I need to upskill on so that I can survive the next layoff round? And so here is kind of the formulation I’ve been sharing with folks, and that is, start with your strengths. You know, grab a piece of paper. What do you have deep inside that you know can make a brittle process smoother using AI, because that’s a sale right there, because where AI is good―like so learning how to use AI right now is easier than ever. There’s more and more courses, there’s more and more videos, there’s more and more ways of using the existing tools. Avail yourselves of these, in fact, and I have a couple resources I can share at the end.
Learning how to build an AI is hard, right? And so take some of the challenging things that are necessary in your discipline, your domain, whether it’s mechanical engineering, industrial engineering, computer software, whether it’s in analytics, mathematics, consultant. And then ask yourselves, okay, how do you make this brittle process better? Where you are essential to making the AI work more effectively, more accurately, etc. And start with that―you want to come from a human centric approach, because the whole point is not to obsolesce yourself, right? Then think, Okay, well, what are the tools out there that I can actually do this and make that proof of concept work in that process? Is it easy? No. Is it hard? Yes, it will it be disappointing where you may just want to, you know, shake your fist in the air, yes, but the same thing I teach my students is this, friction is critical for your development. Friction is how we learn.
The next piece is, you know, how do I encourage people?
So, you know, one thing that we teach here is that, you know, in our lower division classes, we do not allow AI because we want students to learn again, that friction, right? We want them to learn, develop the capacities to understand, for example, computer science from the fundamentals. But then in upper division, we encourage it heavily, so that they can actually see the maximum leverage that they can get from a drive efficiency and drive productivity. But like for those who have been negatively impacted by AI, you know, displacing jobs and work, what I like to ask people is, you know, consider what is so important in your work, like critical, but that’s really hard for computers to replicate? As you get through that, that actually becomes an opportunity.
And I had a meeting with another organization just recently, and they had asked me, like, you know, what’s one of the best tools that you have to help students? And I say it’s this. It’s a notepad. Because with a notepad, I can define an algorithm that needs to be built, I can define a problem space, an unmet need. I can create a whole business from a notepad, you know, and so human creativity is one of the key things that we have that a machine does not. Human resourcefulness is another key thing that we have that the machine does not and in fact, you know, for people who are not familiar with how patents work, one of the key factors to look for in a patent filing is, what was that stroke of genius that this author had when they filed this patent? Being able to emphasize that is actually critical to a patent being successfully awarded, that, I kid you not, is the key. There’s so many unmet needs out there, so many problems that need to be tackled. They’re looking for someone to come along and bring a novel, creative solution to it.
From my vantage point, there’s just thousands and thousands of opportunities that need the right person to address them. So if folks are worried and panicking and anxious, your value as a human being is tremendous, and there’s a God that loves you and so please, you can avail yourself of these opportunities and suddenly turn your entire perspective around.
AI and Higher Education
You’ve expressed excitement to cultivate curiosity in the next generation of Biola students. But do you have any concerns for those who are now growing up in the age of AI? How can students develop an analytical mind capable of producing original creativity, when the answers are often at their fingertips?
Yeah, now that is a clear and present risk, right? So the way that I get at it is I want students to have a moment of direct honesty with themselves. And that’s this: human nature always looks for easier, cheaper, faster, because we, you know, the desire for laziness and comfort and security is a result. It’s not an originating point, but it’s a response. It’s a natural response to the fall, right? So, so that’s the first piece. We cannot be naive. And we have to, you know, encourage students and young people and and all people that we cannot be naive, that just because, you know, if we have such easy inconvenience right away that we won’t completely hand over and outsource our ability to think to it, that’s just human nature. We cannot be naive about that. So that’s one.
The second is then, yes, taking time to double down on uniquely human characteristics, and so I’m gonna, I’m gonna say, I, I believe that the best creativity comes from doubling down on the six inches between our ears and not relying on other tech for that. Yes, you can use other tech to iterate and test and stress test something, but for innate human creativity, no, I think the best things come from humans. And so for that, where I tend to, you know, try and focus students, is think about again, unmet needs again, but then also think about what excites you? Where are your natural affinities?
You know, I started reading science fiction again, and fantasy again, because I forgot what joy it brings to live the life of a dolphin jumping out of the water. And as I’ve been doing that, it’s kind of brought in focus again. Why fantasy? Why creativity? Why imagination is so powerful. And so where I see the need is for students, is that if they don’t hone their capability for creativity and resourcefulness, they can end up in a position where they will just be a glorified data clerk. And essentially, in terms of the divide, I see two divides. You know, one is you’ve outsourced and you haven’t developed your critical reasoning capabilities, and so you become a data clerk. The other side of it is you’ve doubled down on human creativity and resourcefulness and ingenuity, and in that way, now you are a leader creating AI or making AI do something for you and reducing time, energy, cost and leveraging and returning back to you, efficiency, productivity and iterative gains.
And so for that piece, how do you get creative then? You’ve got to read a lot. You have to read a lot, experience and surround yourself with good taste, good, diverse material, art, galleries, the classics, humanities. I mean, there’s a reason why our school is one of the few remaining Great Books programs in the United States. Our students read a tremendous amount of original primary sources. And what happens is, then it provides you a sense of what is good taste. You get more discerning. And you know the way, the colloquial way, I like to say to my students is, don’t grab the McNuggets when you could have steak dine-in. And so it’s, it’s an encouragement. And when students start reading rich, high quality material, creativity is a natural byproduct from that.
Technology and the Church
Lastly, in addition to supporting operational aspects of ministry, how else can the church use technology?
I have such a high view of the pulpit and what it means for Scripture and teaching, I get nervous when I hear of instances where pastors may rely on AI in sermon development and in their teaching. And the reason for that is because, if we hearken back to the Old Testament, you know, why does God call pastors? You’re called to be the anointed, and part of that is being able to listen to the Holy Spirit to speak what the Holy Spirit wants their congregations to hear with high fidelity, high accuracy. And that means you know, as a sports fan, sometimes you have to call an audible, which means you know you were going to preach on this, but then the Holy Spirit is telling you to preach on something else, to convict the hearts in their congregation. And you know, all of us have to every Sunday, right? You know, why do we? Why do we meet for church every week, 52 weeks out of the year? It’s because we forget, you know, we we see in the New Testament the apostle Paul, starting a couple books with remember, remember, because we forget, forget. And with that, we, during the course of a week, we forget to keep our eyes fixed on Jesus.
And so, because I have such a high view for the anointed, you know when preachers sometimes see teaching on Sunday as pulpit supply, that worries me, because it’s not a production. It’s not an assembly line, it’s not a mass produced product. No, it’s you hearing from the Word of God, from God himself, and you being a mouthpiece for us, because we need to hear from God. We need to hear through our pastors and our preachers who we rely upon to help us come back to center line, to be convicted, to feel a knife cutting across our heart, convicting us of the sins that we have accumulated over the course of the week, both intentional and unintentionally, and then having that moment of grace and regeneration. And so I get nervous when I see pastors using AI in their sermon development.
Where pastors could be using it effectively, is in cross referencing and trying to check validate sources across other things, and then going back to primary sources to check those. I think that’s obviously a good thing, especially when you know you’re flipping back and forth with Koine Greek you know, that’s not something that you can just do sometimes. And so in those areas, again, as a nontheologian, that looks like an appropriate use of those technologies.
The other one is in just getting the word out in devotionals and getting the word out in terms of how to reach a secular population that is so in need of a genuine personal connection by, you know, putting out there, hey, this is what―if you’re struggling with, you know, unemployment or anxiety or this, and then sharing a word from Scripture, Scripture never leaves us void, right? And sharing word of Scripture and then explaining that, you know, nonjudgmental, warm, winsome, humble way, the same way Christ did when he came here in person, that will draw people, that will give reassurance, that will continue to help people who are suffering deeply. And so I think the more pastors are producing their own content from their own office, from their own church, from their own desk, you know, writing a quick thing or 30 second video about what this one verse in passage means if you’re suffering from addiction or anti anxiolytic medication or divorce, or, you know, same sex attraction, or any of these things. You know, we should always be the ones who are first to love people. And so I think pastors making content and distributing it far and wide, I think that’s a wonderful way of using technology.
Recommended Resources
- Think Biblically Podcast | AI: Prospect and Perils
- Resources to better understand AI & Policy: