Tech Refactored

S2E20 - Yes, That Laser is Tracking Your Eye Movements

December 17, 2021 Nebraska Governance and Technology Center Season 2 Episode 20
Tech Refactored
S2E20 - Yes, That Laser is Tracking Your Eye Movements
Show Notes Transcript

It’s the 50th episode of Tech Refactored! We’ve talked about so many different technologies - but this is one of the coolest and scariest! On this episode Gus is joined by on of our faculty fellows, Dr. Bonita Sharif. Dr. Sharif is an expert in the use of eye tracking technologies to enhance the efficiency of software programmers in searching through thousands of lines of code.

A practical guide on conducting eye tracking studies in software engineering

Can the ez reader model predict eye movements over code? towards a model of eye movements over source code

Determining Differences in Reading Behavior Between Experts and Novices by Investigating Eye Movement on Source Code Constructs During a Bug Fixing Task

Nebraska Today Article

Disclaimer: This transcript is auto-generated and has not been thoroughly reviewed for completeness or accuracy.

[00:00:00] Gus Herwitz: This is Tech Refactored. I'm your host, Gus Herwitz, the Menard Director of the Nebraska Governance and Technology Center at the University of Nebraska. Today we're joined by Dr. Bonita Sharif, an expert in the use of eye tracking technologies to enhance the efficiency of software programmers In searching through thousands of lines of code, Bonita received her PhD in computer science from Kent State University.

And has since joined the Department of Computer Science and Engineering at the University of [00:01:00] Nebraska. Benita, thank you for joining us. Happy to be here. Gu, can we just start with you telling us a bit about, uh, your current position at the university and your research? 

[00:01:12] Bonita Sharif: Absolutely. I came here in 2018. I am now associate professor in the newly formed school of computing, right? Let's not forget that. Uh, I direct the software engineering research and empirical studies lab here in the school. Um, my research works is mainly, uh, situated at the intersection of. Software engineering and human computer interaction. This is neat because I'm interested in understanding how software developers work and, um, one aspect of it is the interaction they have with the system as well as other people.

So software engineering is a, um, subfield. In computer science, which is mainly focused on building systems all the way, starting from [00:02:00] requirements to when we deploy it and maintaining the system over time. 

[00:02:04] Gus Herwitz: So you mentioned a couple of things there. Software engineering as a subfield of computer science.

And also you, uh, noted as I should have noted in our introduction that here at UNL, the Department of Computer Science and Engineering has recently been reorganize. Into a school of computing. Software engineering has the word engineering in it, and computer engineering has often been how the field of computer science or part of the field of computer science has been conceptualized is software engineering.

Uh, this is gonna be a hard question. I expect I'm not gonna go anywhere good with this. I- is software engineering, an engineering field, a computer science? What, what is software engineering? 

[00:02:48] Bonita Sharif: Yes, definitely. Software engineering is an engineering discipline. Now, there are people out there that will fight you on this, okay?

But it is engineering because now if you do, okay, let me [00:03:00] rephrase that. If you do software correctly in the correct way, it is an engineering discipline. The problem is most people don't, and this is why we have software in this data, this today. The reason why it is engineering, there's a disciplined way of going.

Building software. There is a method to this, right? We, we document things as we go, and if you do it correctly, now, again, most people don't do that. Most developers don't document. If you see 90% of the cases, they don't do that, but. If your firm does not do that, I would say either evangelize, convince them to do it or leave and join another firm that does that.

So, uh, yes, it's definitely engineering. There's a, it's, it's software. You can't compare it with building bridges, right? You can actually see a bridge. You cannot really see software, and that's what makes it hard to, to actually see what's, what's working and the intricate details that happen with the different systems.

Integrated with each other to [00:04:00] just build something that you see on a phone. That's all you see as a user, right? The, the interface. There's so much more behind the scenes. A good example of this is Google, right? There's so much, uh, complexity in, in the search, but all we see is a. 

[00:04:16] Gus Herwitz: Xbox. Yeah. So at some level the, the engineering component of software engineering is even more important from the discipline and actually building this stuff perspective than it is with physical things like bridges because no one can see and it's, it's easier to take shortcuts and harder to understand intuitively.

So you, you need to be a more rigorous engineer and approach these problems from an engineering 

[00:04:42] Bonita Sharif: perspective. Yes, a hundred percent. And this is why, because of this invisible nature of it, we need to, and we follow this process in our, both, in my lab as well as in the software engineering degree program here we have at this school.

Uh, we make sure we document everything the [00:05:00] workflows are in place, because if we don't have that, we cannot possibly remember everything, every little detail. So everything needs to be 

[00:05:07] Gus Herwitz: documented. So you, uh, had started also by saying your work is at the intersection of software engineering and HCI, or human computer interaction.

Can you tell us a bit more about, uh, HCI? 

[00:05:20] Bonita Sharif: I'm interested in understanding what, for instance, I'll give you one example. I'm interested in understanding how, let's say a expert developer is different from a novice developer. And in order to do this, we use certain objective methods such as, you know, uh, biometric methods, such as one of them is eye tracking.

And we place these, uh, we, we place these objects in, in a way such that we can track. They're looking at, and this, this interaction is useful to understand how, let's say developers fix bug. Now we could do that without that, without the interac, the, uh, eye tracking, right. [00:06:00] But we would not possibly know, we, we would get the end result, how they fixed it, but we wouldn't know what they did in the process.

And that interaction, I think, is super important to understand. How we can help a novice, let's say improve and become an expert eventually faster. 

[00:06:21] Gus Herwitz: So I, I love that your work is kind of meta in computer engineering and software engineering because you're using computer science research to figure out how to improve the state of software engineering, um, which is itself a software and computer engineering discipline that that's a perfectly meta.

In the traditional meaning of the word, not the new Facebook branded version of the word. 

[00:06:45] Bonita Sharif: You hit the nail on the head there. Yes.  

[00:06:48] Gus Herwitz: Uh, so how, how did you come to, uh, focus on eye tracking? 

[00:06:54] Bonita Sharif: This is by complete accident. I was a grad student back at Kent State University. I was [00:07:00] working in a completely different area, software traceability again in software engineering.

My advisor asked me to check out this eye tracker we had in the library. This was a not as good as the trackers I have now, but it was something that we had there and he suggested, you know, you can replicate this study, possibly think about replicating this. And I was like, Why is he asking me this? I'm not sure.

So I went there. You know, as a grad student, you do what your advisor tells you to. I have a great advisor, by the way. So I went there and I checked out the, the place and, and since it was not my tracker, like I had to kind of get, get time to use it, it was in the library. I replicated a study. Again, this was just for.

Point this out. Uh, I had no idea that I would be at this position actually doing this for a living right now. So I went to the library, set up this, uh, experiment, and the, the goal of the study was to understand what's better in terms of identifier style. Uh, Camel [00:08:00] case or underscore. Okay. So the study was purely to see which, which one's better, which one takes longer to read.

And that's what I did. Uh, fast forward 10 years, this is actually my most cited paper now, and quite controversial actually, . 

[00:08:15] Gus Herwitz: So I, I have to ask, uh, what, where's the controversy? How's it controversial? 

[00:08:20] Bonita Sharif: So identifier styles, these are like religion for some developers they stick with, You know, I only use one or I only use the other.

I say this because there was a Reddit post about this paper actually a while back. I can find it. It's still there. Where. People were basically arguing with, with, with each other about the results of this study. One half was towards Camel case and the other half was towards underscore, and they were just fighting with each other about why one is better than the other.

Uh, using this paper as either evidence or saying that, you know, Oh, it doesn't matter, like. You can pick your side. Right. It was funny to [00:09:00] actually, not funny, quite intriguing for me as an author of that paper to read these things, so there was quite a controversy. Stir up on that. 

[00:09:09] Gus Herwitz: Can you just explain, uh uh, for listeners, we're talking about in computer coding, when you're writing computer code, What names you give to things like variables and functions, and there are different ways of doing this, like, I guess I'm about to explain this and I, I shouldn't, I should ask you to.

Okay. Can you explain what the difference between Camel case and underscore is and what, why it matters? 

[00:09:31] Bonita Sharif: Absolutely. So take a compound word, for example. Take the word total cost. Okay. When you write that in Camel case, you will basically spell it out as T O T A L, all lowercase. You would capitalize the C and then have lowercase o s t.

Okay. But only c capitalized. If you write the same compound word in underscore, it would be total underscore cost, everything lowercase. Now, there are some issues with [00:10:00] readability with both those things, right? Uh, and some people prefer one or the other. 

[00:10:06] Gus Herwitz: And th this is actually bring us in a sense, to the human computer, uh, interface sort of idea.

The, the reason this all matters. For computers, when a computer's reading through and uh, uh, processing computer code, a space will break things up. So if you want to have a variable that is meaningful, the total cost of whatever function that you're doing, the the total cost, uh, of items in a shopping cart, you can't have it be total space cost.

You need to have it all just be no spaces. So how do humans. Interface with the way that computers are going to be reading. The variable name there is, is that basically a nice dovetail between the topics? Yes. I, I, I love this story that you tell, uh, your, your advisor suggested, why don't you go do this, uh, a project and it took you down this path.[00:11:00] 

True. And thi this is what academics in many ways should be. It's really hard to know when you're a grad student or even when you're a, a, a professor at any point what direction, uh, your career is going to go. Because let the research take you in interesting places. If it's interesting to you, it's probably.

Interesting to others if you can do, uh, tractable work around the topic. So can you tell us a bit more about what eye tracking is and how this research works?

[00:11:29] Bonita Sharif: Yes. So the eye tracker itself is quite simple. It is a hardware device with a light source and some cameras, sometimes one or two cameras. Uh, one or two light sources it and it basically tracks your eyes and tells you a- a point in space where you're looking, let's call that point. It has an x y coordinate, let's say, and it determines this by shining those, uh, light, that light source towards your eyes, determining where your pupil is, [00:12:00] and then looking at where the corneal reflections are just a fancy word for glint that you see on your eyes.

And then, It basically uses straightforward geometry to determine the x Y coordinate on the screen, and then that's that coordinate or that point rather tells you what you're looking at. That's pretty much what the eye tracker gives you. It doesn't tell you what that point maps to. That is the next step in post-processing that you need to figure that out later.

But the tracker itself just gives you this point in space that you're. So that, 

[00:12:34] Gus Herwitz: that's, uh, really cool. So the, you're basically using cameras to look at the eyes and you're taking what's reflecting off the eyes. It's not, uh, the camera isn't scanning what you're looking at, or it isn't on the screen seeing, Oh, you're looking at this point on the screen.

It, it's, this is applied to trigonometry. 

[00:12:53] Bonita Sharif: You can think of it like that because, um, it, it all, it depends on what you wanna do with that point later on, right? You [00:13:00] can write your own software to do what you want with it, but the tracker itself really doesn't give you anything more than the 

[00:13:06] Gus Herwitz: point. So what, what are you using eye tracking to do and, uh, what, what sort of research are you doing with it?

[00:13:15] Bonita Sharif: So right now we are using eye tracking to understand how software developers fix bugs, uh, how they summarize code, how they basically navigate to understand code, because we want to determine, My goal eventually is to come up with these theories. I don't have a theory right now because we don't have enough data.

To distinguish, let's say, novices from experts. But I would, or even how novices learn over time. But the more data we collect, the the better we can build these theories to understand how developers truly understand co how do they fix bugs? How do they go about what's their thought process in doing so?

And that I think is extremely intriguing. 

[00:13:59] Gus Herwitz: [00:14:00] So we, we could think of this, and correct me if I'm wrong, but, uh, to put, uh, a metaphor on this, this is like reading a recipe. So you could, you could have an experienced cook, uh, you could have a novice cook. You give them a, a recipe. And you want to know how do they read this?

Where do they get confused? Do they understand, uh uh, one lowercase tea versus one uppercase tea as teaspoon versus tablespoon? Or should you spell out teaspoon or tablespoon? Um, is it more useful to have all the ingredients listed as a block up at the top, or do you list them? The first time that they're used in the recipe.

Uh, so you give novice cooks, uh, and, uh, experienced cooks the same recipe, and you literally watch their eyes as they're reading this and figure out, okay, where are they getting confused? Where are they making mistakes? Where are they trying to figure out, uh, what I'm supposed to be doing while the chicken is burning on the, uh, in the oven

Um, is that kind of what you're doing? 

[00:14:59] Bonita Sharif: Uh, yes. [00:15:00] I'll give you another analogy. I just, uh, thought of another one. So think about how people, I play the piano myself and, um, think about how people play the piano, right? Um, a, a novice versus an expert. So what they found out when they use eye tracking while people are playing, they notice that the experts tend to do a look ahead, so they look.

Few notes before they actually play it. That tells the, tells us that they are actually using different strategies. Another concept is they don't look at their hands as much, whereas the novices don't look ahead and they always look up and down at the notes and their hands. So this tells us that we need to figure out a way to.

Help the novices learn this look ahead skill. The same with code, and the same with the recipe idea too. It's, yes. How do we learn from these, these sticking points that are making the novices, let's say take longer in in doing a task. And this is not just for novices. It could be also for experts that are exposed to [00:16:00] an unfamiliar 

[00:16:00] Gus Herwitz: task.

So what does eye tracking add to our understanding of how to improve learning? How to improve code readability, coding, software engineering generally, or I guess just more generally, what, what are the main contributions and advancements that this work, uh, is leading to? Yes. 

[00:16:21] Bonita Sharif: The biggest thing, I see it as a glass box, right?

You can actually understand and see the thought processes of someone as they're doing the task versus after the fact. There are two things that, uh, that occur when you ask somebody what they did after the fact without actually tracking what they did. While they're doing it, one, they. Misreport what they did or think they did.

And two, they, they might just forget what they did while they're doing the task. Right. And I know this because we actually play back gay's behavior to, uh, our, uh, study subjects. And they tell. Tell us things like, I had no idea I was [00:17:00] looking at this thing at this point. I'm not sure why I was looking there, but they were right.

So this can help us uncover a lot of issues in, in design, in the way the code is built, as well as other usability issues in actually using the code. Both developers as well as users can, can be impacted with with this 

[00:17:19] Gus Herwitz: research. So we are talking with, uh, Bonita Sharif at the University of Nebraska, um, about her research into eye tracking and software engineering.

We will be back in a couple of moments to continue our discussion. 

[00:17:35] Morgan Armstrong: I'm Morgan Armstrong, a soon fellow at the Nebraska Governance and Technology Center and part of the Space Cyber and Telecommunications Law Program at the University of Nebraska. Did you know the University of Nebraska College of Law also has a space, cyber and telecommunications law program that started in 2008?

The program features tracks for law students and advanced degrees for established attorneys interested in satellites, international law, [00:18:00] radio spectrum, or just about anything in the great expanse of space. Check them out on Twitter at Space Cyber Law. Now back to this episode of Tech Refactored.

[00:18:13] Gus Herwitz: Hey, we are coming back now for some more discussion about eye tracking, software engineering and possibly, uh, what it's like to gaze through the windows into someone's soul. So, uh, Bon. I'm gonna start on that last, uh, point, the way that you were just describing this, uh, the, the work that you're doing. I, it, it sounds really powerful, really being able to look into and study someone's thought processes as they're playing the piano, cooking a recipe, reading a book, or a, some text or something like that.

What I, I'll, I'll just let you respond to that, that just seems really powerful to me that you can actually kind of peer into individual's thought processes in real time.

[00:18:58] Bonita Sharif: Yes. So let me start [00:19:00] that by, uh, telling you about a actual pain that I had when I, when I first started in tracking and why, and how I resolved it.

So when I, um, first started eye tracking, my goal as a software engineer was to study developers, but when I looked at the software available, I couldn't find anything that let me conduct a study in a way that was realistic enough so all the state of the art systems out there would just let me show a picture or a snippet of code that fit on the screen without interacting it, with it, without moving it.

And this was fine to study small. Code snippets, but you cannot realistically take those results and extrapolate them to real software that impacts real production software that people develop in the field. And this was an issue I had and and I, there was nothing out there, so I decided to build my own.

And we call it now, it's community [00:20:00] infrastructure. It's called iTrace i-trace.org. You can look it up on the website. We are just, uh, coming up for a final release of this in January. It is fully functional. It actually works, It is used by several of software engineering researchers in our community, and we are constantly getting, um, requests to update it.

So let me explain what's unique about it. So if you remember before I said the tracker just gives you the x y coordinate that you look at on the screen now, it'll do that so long as the screen does not. But as developers, we don't work with 10 lines of code. We work with many, many, many thousands of lines of code, and we switch between files.

You know, we search the web, we, we look at stack overflow, right? Who doesn't? So I wanted to support all of these different use cases, and this is why I created iTrace along with my team. It lets you implicitly collect all [00:21:00] of this data. Behind the scenes while the developer works. And so it completely revolutionizes the way you conduct eye tracking studies in software engineering because it is very, very, very realistic.

Just like you couldn't get more realistic than that. This is how software is built in industry and we are basically tracking it in that way. 

[00:21:24] Gus Herwitz: So ba basically, if I were to. Be reading something and thoughtfully look up at the ceiling for a couple of minutes, uh, and ponder what I was thinking. You'd be able to track that as a thing that I was doing.

Whereas, uh, a previous generations, uh, I would have to be looking at a static code snippet or something that, uh, whomever was doing the study had wanted me to study. And that that's just a completely artificial, uh, way to actually, uh, uh, 

[00:21:56] Bonita Sharif: Actually not quite. Let me uh, rephrase that. So, [00:22:00] uh, both systems would actually let you see what you, if you looked off of the screen, what, how ours is different is that with I trace, you can actually.

Switch between files. So let's say you was looking at one file and then you wanted to reference another file that this file was referencing. So you, you can click another file and then our system seamlessly detects that you switched files. You can open up a web browser and look at stack overflow. It will seamlessly switch to that and tell you, Oh, you looked at this comment in stock overflow, or this particular code snippet in a bug report.

Previous state of the art systems, Actually right now, the ones that exist in the market also do not let you do that. It has to be fixed on the screen. 

[00:22:47] Gus Herwitz: So what's the, uh, level of granularity that you're able to detect eye movements at? I, I've seen previous, uh, Technologies that can kind of say you're looking at the upper right corner of the screen or in the [00:23:00] middle of the screen.

That's that've been used, for instance, for, uh, webpage design. How do, how do a user tend to scan their eyes across the page? And that's pretty coarse. It sounds like you're able to do, uh, much finer grained positioning on the. 

[00:23:16] Bonita Sharif: Yes. So we typically look at the word level. Uh, we can look at the character level, but that's not really useful to us.

So we look at the word level, uh, definitely the line level. Sometimes we chunk things together, but yeah, we can go up until up to the word level. And, and it also depends on the accuracy and the quality of the tracker. Right. So you wanna use a tracker that is a good quality that has, uh, higher speed so you don't miss too many things in between.

So there are a lot of factors there as well. 

[00:23:47] Gus Herwitz: I putting on my lawyer and technology regulation hat, I have to ask, do you have any concerns about the technology or how it might be used and how does that affect how you've [00:24:00] approached this work? 

[00:24:02] Bonita Sharif: Yes, there are privacy concerns, but there are ways to get around it.

So for instance, if you are not tracking the I itself and just the data that's coming out of it, and then for also not tracking what they're looking at. So for instance, if you were a company, a software company, and you wanted. To give all your developers an eye tracker, maybe then there has to be some policy in place or your, some trust or some kind of system in place that your developers feel, you know, comfortable in letting you use this to track them.

For instance, when I work, I certainly do not code. You know, four hours at a time. I have 20 minute increments, and even within those 20 minutes, I don't always just look at code. I might go off on a tangent, look at a website maybe. Now if that happens, instead of recording what the website is, you can just say, Oh, this was off task, rather than saying what it [00:25:00] was.

So there are ways that we can protect privacy by just not recording certain things that we don't need to record. 

[00:25:07] Gus Herwitz: And in, in terms of the physical setup of this equipment, this isn't the sort of thing that evil corporation X, Y, Z could just enable on someone's laptop, make sure this person has read this term of service.

This is actual, you, you have a, a lab and you need to be in a facility that's using this equipment. So, uh, it couldn't be surly used. 

[00:25:32] Bonita Sharif: Well, right now there is a push to use webcams as trackers. Let me tell you that they do not work well, and some of these results scare me because they're making conclusions that are simply not accurate.

We have actually done a test in the lab to compare webcams used as eye trackers with an actual tracker, and there was 5% overlap. We haven't published this yet. We are planning to eventually, but a webcam is no substitute for a tracker. So, [00:26:00] you know, I would be very, very uncomfortable with if somebody used a webcam and said, Oh, you didn't look at this, which they could do, but we need to be careful about that.

[00:26:08] Gus Herwitz: So, uh, for those listening at home, you would not have just seen my facial response, uh, when Bonita mentioned that it was a, a wow, I, I shocked that it's only 5% overlap. Uh, that that's really scary. And, uh, yeah, I, I'm sure that we're starting to see, why am I say, sure. I know that we are starting to see these technologies in the wild and companies trying to, uh, really promise these capabilities.

So please get that paper published. That's, uh, really important work. What, what do you, is there anything that you wish people, uh, understood about this technology? 

[00:26:47] Bonita Sharif: Yes, I definitely wanna point out that it's a lot of fun to use, but it does take a lot of time and effort to actually not just conduct a tracking study, but also to design it.

We spend [00:27:00] about, nearly about half a year, six. Four to six months. You know, designing, piloting, making sure the study is correct, the tasks are appropriate. It's not plug and play like the vendors make it out to be. Okay? So it's, yes, you can plug and play it, but then what next? Right? You have to have some functional use case to using it.

And also think about the realistic setting when you are conducting studies. Like, who is your end user? Who are you studying? If you, you know, want to study, let's. Java developers then don't focus on people who don't know Java, right? It completely throws off your your results. So just be mindful about these things and then not everything needs eye tracking.

I do a lot of stuff that does not involve eye tracking. There has to be a reason why you want to use it. 

[00:27:51] Gus Herwitz: What about for those, uh, who do want to use it? You mentioned I trace and, uh, that this is a community infrastructure. What [00:28:00] sort of researchers might be interested in using this? Um, what capabilities, capacities do they, uh, need to have to do so?

And um, again, I'm thinking, uh, my own legal research, most legal. Probably don't have the capabilities to implement iTrace. What if there was someone who wanted to understand, uh, how users read terms of service on a website and legal enforceability that we should or shouldn't give to certain terms? What, what advice would you give them if they wanted to start working, uh, uh, with these sort of tools?

[00:28:35] Bonita Sharif: Yeah, so iTrace started with software development in mind, but it's actually very, uh, useful to other, uh, disciplines as well. And let me explain. So iTrace is community eye tracking infrastructure, uh, but it is set up in a way that it's very modular. So we have a core system and then we also have these plugins that connect to it.

So we have, one of them is the, i-Trace Chrome plugin. So [00:29:00] it basically enables eye tracking within Chrome, which is a web browser. Now you can, and because of this, you can have, let's say, uh, you know, terms of service open and you can have people read it and see whether they truly understand what terms of service are, or you could have legal documents in the browser and, and then record at the word level.

To see what they're reading. You can detect reading speed, for instance. You know what, they skip regressions, things like that. So that could also help you to better write the documents, right? There are so many other possibilities. And then other things such as Excel. Think of Excel, right? We could write a plugin.

Eye trace is set up in a way that you can write multiple plugins, so someone could write a plugin for Excel that detects how people work with formulas or, or code within Excel. That's an option. We also support Visual Studio, which is a, uh, development environment for developers, and then we also support Eclipse and Atom.

So we support quite a [00:30:00] few now and then it's extensible. If you go to our website, you can learn more about how you can extend it. So we definitely want people to use our core and extend it. 

[00:30:12] Gus Herwitz: What's on the horizon when your, I, if we were to track your eyes and your gaze is looking out to the future , um, what, what do you see both with your own research and the direction that this technology generally is going?

[00:30:26] Bonita Sharif: If you look at the market and where people are investing things, you will see that a lot of money is being put into the eye tracking field. Apple, Facebook, they're all looking for. That killer application. We don't have that yet. We need a killer consumer application to have eye tracking takeoff. Right. We don't have that yet, so I see.

But I predict 10 years from now, every laptop will have an eye tracker. Maybe not 10, maybe 10, 15 years from now, but it will eventually happen. And when that happens, how are we going to use [00:31:00] it? Right. I know how I would like to use it to help me understand my workflow as a developer. It being self-aware of how you work just helps you, helps you with your productivity, helps you just plan your time, even makes you learn things about how you-

let's say, go about trying to fix bugs, for instance. This is a, a learned skill, right, Even for yourself. Uh, and then it can help others as well. For instance, you can show these, these patterns to people who are onboarded on your project, for instance. It can reduce the time for people that come in to your system.

I mean, imagine a system where you could actually. Show people the eye gaze of how you would, let's say, navigate this project instead of having the person right there. It will also enable a lot of other things to be reused in the system. So, uh, we are not there yet, but I think, you know, we have to resolve some things around.

There is some policy issues that need to be resolved with ethics and privacy, but [00:32:00] I think there is a way to get around that if we collect only what we need and not necessarily everything we. 

[00:32:07] Gus Herwitz: And I, I just have to put a, a shout out in there. Uh, one of, one of the great things about, uh, Bonita that I'm happy about is that she's one of our fellows at the Governance and Technology Center, uh, here at the university where we think about and talk about, uh, these ethics, tech ethics, privacy regulation sort of concerns.

Because I expect many listeners are just imagining all the scary "well my employer's going to be able to manage every single thing that I, I do. Facebook is going to know what posts I'm not just clicking on, but looking at" concerns there, it's, uh, important and, uh, good to know that, uh, these are things, uh, that are in the discussion now and.

I gotta say I have my own dream application for this technology, which is incredibly simple. I have too many monitors [00:33:00] connected to my main computer at home, and I would love it if I could just, I, I often lose track of the, my mouse pointer on the screens. I would love it if I could just push a button and have the mouse pointer appear wherever I'm.

So I'll, I'll put in my plug for that as a future application. 

[00:33:17] Bonita Sharif: Oh, that's- that's easy to do, Gus. We'll do that for you. No problem. 

[00:33:20] Gus Herwitz: Okay. Well that, that's what I'd like to hear. Any last thoughts that you want to, uh, leave listeners with? 

[00:33:27] Bonita Sharif: Yes, I wanna say that eye tracking is such a powerful technology, but at the same time it can be misused.

So, um, use it correctly and it'll give you these unique insights as to, you know, how people work. And hopefully in the correct hands, it can have a lot of impact and improving technology and just in general life and software that helps us live a good life, I guess. 

[00:33:52] Gus Herwitz: Okay. Well, thank you, uh, Bonita for this discussion.

Really fascinating work. And I, I'm just, uh, I, [00:34:00] one of the things, uh, uh, I know I have colleagues who are interested and excited about, uh, collaborating with you on some research. Uh, so look forward to continuing the discussion. So thank you and thank you to our listeners. I've been, Gus Herwitz. Thank you for joining us on this episode of Tech Refactored.

If you want to learn more about what we're doing here at the Nebraska Governance and Technology Center, or submit an idea for a future episode, you can go to our website at ngtc.unl.edu, or you can follow us on Twitter at UNL underscore NGTC. If you enjoy this show, please don't forget to leave us a rating and review wherever you listen to your podcasts.

Our show is produced by Elsbeth Magilton and Lysandra Marquez and Colin McCarthy created and recorded our theme music. This podcast is part of the Menard Governance and Technology Programming Series. Until next time, keep your eye on the ball.[00:35:00]