Aaron Gustafson -- PART 2
===

~Testing, testing, testing.~

~Test, test, test.~

~Yeah, I don't remember how exactly I started the story, so I'll try, try my best. Um, yeah. All right. So I was doing some, uh, consulting for a large retailer here in the US and I was talking to them about, Devices that they were testing on and, and sort of their, their test matrix, their device lab, that sort of stuff.~

~And they walked me through what devices they were using and it was kind of what I expected. They had a bunch of iPhones. They had a couple of iPads. Um, I think at the time the iPad mini had just come out, so they had a couple of those that they'd added. Um, and then they're like, we have a couple of Android phones as well.~

~And I was like, okay, great. Now they sold very low end. Um, Tablets and they had very low specs and not great displays. And you know, I'm, I'm imagining not a very, you know, not the fastest wifi chip set and so on and so forth. Not the most powerful processor. Um, I think they ran about $50, $75, something like that.~

~And I was like, how many of the devices in your device lab or devices that you sell? And I was shocked because there, like you could have heard a pin drop. Nobody had actually considered the fact that somebody who comes into their store purchases a, you know, device from them and then maybe needs to. You know, use their services on their website and expects it to work.~

~Um, and so I think, you know, we, we tend to, yeah, just kind of be creatures of the environment that we're in. And if, if we don't work hard and intentionally to expand the sphere of influence, um, the sphere of our influence of what, what is influencing us, um, it makes us really, it makes it really challenging for us to build products that are actually gonna work for our customers.~

~Um, you know, I think WhatsApp is another great example of this. You know, the, when WhatsApp started, I don't even remember now, it's some astronomical number in, in the iOS app store. I think there were something like 1500, but there's probably more like 3,500 different chat apps. Um, you know, so many different options for people to try.~

~Um, But WhatsApp with a very small team, you know, not only went to iOS, they went to Android, they went to Symbian, they went to Blackberry. I think they went to a bunch of different platforms, right? And that's how they built the sizeable customer base that they had, which is why they got acquired for so much money.~

~Um, But I think often as developers, when when we go in creating a, you know, a new Blue Sky app, we think, oh, we gotta be in the app store and we ignore everything else that's happening globally and, and even domestically, wherever, wherever our company is based, we, you know, in the, in the US a lot of us on the coasts overlook what the experiences are in the middle of the country, in, you know, rural areas or something like that where people are not on.~

~5G or don't have access to the latest and greatest devices. You know, we focus on people who, you know, are at the higher end of the income spectrum, ignoring the fact that people who fall into the 30,000 and under, uh, category in the census actually are about half the people in the us. Um, so we're really just.~

~Artificially suppressing the potential reach that we have for our, our customers and the money that we could be making, um, or the, the people that we could be providing services to. And certainly that is a loss for companies, uh, in terms of potential lost revenue. But it's also a risk for companies and organizations that are.~

~Basically held to account, you know, if you're, if you're thinking about, um, any sort of governmental programs, if you are a utility, um, if you're a bank, you know, these are, these are services that are necessary. And if you are cutting off your customers because they don't have, you know, the devices that your developer team has, that's a problem.~

~Right. So that's why progressive enhancement continues to be important,~

~Right, exactly. So I'm, I'm just curious, um, I guess like from what have you seen, like how does it, when it comes to like this accessibility question, it's like I.~ [00:00:00] How do you know when something is too old to support? ~Like, like where's the line where like, you,~ obviously you can't support every single device that has ever existed, but ~like~ from like your, from what you've seen, ~like,~ like how do you come up with that decision where ~like,~ okay, this is probably as far back as we can go, ~I guess.~

~Yeah, I, so~ I'd say on the whole. The cutoff is not as hard as. ~Like,~ it's not as hard and fast. Like we absolutely cannot support these devices. The only place that you run into that is with, ~um,~ encryption. So older devices are not gonna have newer forms of encryption that maybe~ the,~ the server that you happen to be running on only allows, or there's, ~you know,~ Problems security-wise with serving, ~you know,~ a certificate that would work on an older device.

~So I feel like that~ that's one area where you can actually, and you should have a hard cutoff and it's unfortunate, ~um,~ but maybe you have, ~you know,~ some sort of experience that could be, ~um,~ H G D P with some, ~you know,~ so basically not putting things like your FAQ and how to contact customer service and things like that behind SSL so that you make sure that somebody can still access [00:01:00] those on a device that doesn't have.

You know the latest and greatest security packages. ~Um,~ But beyond that, I really like the, ~uh,~ approach ~that,~ that, ~um,~ Brad Frost proposed, which was having this concept of support versus optimization. So looking to support as many devices as possible. So ~that's,~ that's the long tail ~of,~ of devices, and that's where having that functional baseline is really useful.

And then you're focusing on optimizing only for. The devices that you're, ~you know,~ that you are seeing used and maybe a handful of ones that you expect your customers are using or that you've, ~uh,~ seen your customers use, ~um,~ in access logs ~or,~ or something of that nature. ~Um,~ I would throw in there making sure that you are actually checking, ~um, ~usage, not requiring JavaScript because a lot of people rely on their analytics package, ~um,~ which is JavaScript based to collect the device data and browser data.

You wanna make sure you're actually checking the raw logs, the raw access logs as well, to be able to get a clear picture of who is accessing. Because if there's not JavaScript running, your JavaScript package isn't [00:02:00] gonna collect any analytics. ~Um,~ so I've had people say to me, ~well,~ we don't have any users that don't have JavaScript.

And I was like, okay, ~well ~how are you recording it? ~Well,~ we're using Google Analytics. Which is JavaScript based. So yeah, that's obvious. You wouldn't have, you wouldn't have those show up. ~Um,~ so you need some fallbacks, right? You need ~to,~ to make sure you're getting ~the,~ the complete picture. ~But, you know,~ that allows your team to, ~you know,~ not have to spend a lot of time.

Focusing on older devices. ~You know,~ maybe you bust out, ~you know, an~ an old device that you picked up on eBay that you know, still has wifi, so you don't have to have ~like~ a data plan for it, but just to like spot check something to see if it works. But you don't have to over index on that and you can spend your time investing in the future.

And I, I think when you. When you adopt the progressive enhancement mindset, ~it,~ it takes a little bit of time to shift over to that, ~to,~ to ~kind of~ shift to that sort of layered thinking. ~Um,~ but once you do, it becomes really easy to support those older devices because ~you're,~ you're already thinking about, okay, what could happen?

What's, what are my dependencies [00:03:00] that could not be met, and how do I ensure that there is an experience for users that, ~you know, don't meet, you know,~ don't have that dependency? ~Um,~ In other words, it's really about removing fragility from our code. It's about building more robust applications. ~Um,~ and when we think about it that way, we do get to focus on those new devices, focus on those new APIs as long as we are not, ~you know,~ cutting off people artificially.

~Um,~ at the same time, I do think there are instances where I. Let's say you're building an image editor or a video editor online. Yeah. ~You,~ you probably need to have some a, some APIs that are non-negotiable. And in that case, yes, by all means, that's where you should be investing your time and you don't need to worry about having experience for ~older devi~ older devices, but, Your marketing site and such should be able to be used on those older devices, ~you know,~ regardless.

~Um,~ and you need to have ~clear, um, you know,~ clear guidelines as to ~like,~ here's [00:04:00] what we're supporting, here's how far back we're supporting. And it's helpful to tell people why as well, that it's not like just some arbitrary thing, like we need a browser that supports, ~you know, in,~ in the case of us recording this online,~ um, you know,~ we need microphone support and in this case video support as well.

~Um,~ So if we don't provide access to that, ~you know,~ you don't get to have an experience. ~You know,~ same thing would happen with Teams or Zoom ~or,~ or what have you as well, right? ~Um,~ there's a certain bar that needs to be met for some applications, but I think a lot of times as developers we are inclined to, to like.

Make our lives easier by only having to test certain devices and just setting arti our arbitrary cutoffs, ~um,~ to try and make our lives easier when we should be thinking about how can we support the widest range of potential customers in order ~to,~ to build value for our business.

~Exactly. That's awesome answer. ~Yeah, I think ~like ~one like attitude I see sometimes is, ~well,~ we don't have the resources to test all these things, ~so.~ We're just gonna test what I have in front of me and if it works there, that's good enough. ~Right. Um,~ so I wanna shift this topic to you. I was wondering if you can ~kind of~ tell [00:05:00] us about your role at Microsoft and what you're doing there.

~Yeah.~ So now I have shifted, as I mentioned, over to focusing on accessibility innovation. ~Um,~ so I am running the AI for Accessibility Grant program, which just wrapped up its fifth year. Through that program, we have funded, ~um,~ all kinds of ~different pro uh,~ different projects and programs. ~Um,~ some of those have been things like, ~uh,~ mentor, which is a, ~um,~ It's, ~uh,~ like a job placement, ~um,~ network basically for people who are neurodivergent.

~Um,~ and it matches employers who are looking for neuro neurodivergent talent with neurodivergent folks who are looking for work. ~And it's,~ it's a really cool, ~um,~ A really cool approach to job seeking. And it's interesting because the way that they work it is you build a profile on there as a job seeker, and you say what accommodations you need, let's say ~a,~ a quiet environment.

I need to not be, not have fluorescent lights, overhead, ~whatever,~ whatever it is that you need to be able [00:06:00] to be your most productive. You identify what your. ~Like~ awesome skills are, like, which things you are particularly good at. ~Um, you know,~ whether that's, ~you know, I'm,~ I'm really excellent at, ~you know,~ debugging code or I'm really ~excellent,~ excellent at, ~you know,~ doing really focused deep work.

~Um, and, you know,~ things that are somewhat negotiable for you from a, ~um,~ A work area, ~um, you know,~ work preferences, that sort of thing. Like I, I like to work flexibly. ~Um,~ those sorts of things. And you put that together in your profile. And then, ~uh,~ job posters will say what they can accommodate, ~what,~ what accommodations they have built into their offices and to their work styles, that sort of stuff.

And they're using AI to match those. Profiles and actually suggesting people to the potential employer that, ~you know,~ here's some profiles that match, here's the percentage that they match based on what it is that you're able to do in terms of accommodations, in terms of what you need, of skills, et cetera, et cetera.

~And then the, uh,~ potential employers actually then reach out to the prospective employees, ~um,~ in [00:07:00] order to begin that communication. So it's not something where the job seeker is having to, ~um,~ Spend a lot of emotional energy, ~you know,~ tracking down and applying for, and, ~you know, given,~ given today's climate, ~you know,~ getting rejected from jobs or not gotten back to at all, et cetera.

~Um, And it's,~ it's ~sort of~ flipping the script on a lot of ~the,~ the approach to, ~um,~ to job placement. And I think that's really cool. The entire team behind the product is neurodivergent, which is pretty awesome. ~Um,~ and so ~they're,~ they're really approaching the whole thing with the mindset of ~what,~ what it is that they need as a community.

~Um,~ so that one's super exciting.~ That was funded, uh, before I joined. Um, But, you know,~ some other things that ~are,~ are happening in that space. We're doing some work right now with, ~um, a,~ a project called I Will in India, and they are trying to address the need for mental health services in India where they only have, I don't know if I'm gonna remember the stat perfectly, but I think it's ~like~ 0.4 mental health practitioners per hundred thousand people.

~So they're,~ they're,~ you know,~ Really don't have the number of professionals that they need to serve ~the,~ the population that they have. And ~so,~ What the I Will Project is doing is [00:08:00] training a, ~um,~ chat bot end-to-end in Hindi to be able to have cognitive behavioral therapy sessions with, ~um, the,~ the people that need it.

~Um,~ and ~the,~ the organization that. Pitched us on this project already had a successful C B T chatbot in India in English, but they were looking to branch out and be able to support more people in Hindi. And it was really important to us in our discussions with them ~that~ that model be trained end to end in Hindi so that you weren't doing leapfrogs from Hindi to English and having, ~you know, the,~ the chatbot figure out what to say back in English.

And then translating it back to Hindi where, ~you know,~ we're dealing with mental health here. This is a really. Important thing to get right. ~So, you know,~ we wanna make sure ~that ~that model is being trained correctly. We also wanna make sure that. There's not ~a,~ a creep in of additional Western influence to the cognitive behavioral therapy that's taking place.

~Um,~ and so that was another important aspect of this and why training that model end to end in Hindi was important to [00:09:00] us. ~Um,~ so lots ~of~ of really cool projects like that, ~uh,~ that are taking place. ~Um,~ more that I can't talk about yet, ~but, ~but just some really interesting things where we're, people are just using technology to really.

Open up new opportunities for people with disabilities in their employment, in their education. Certainly. ~Um,~ like MBI is another one that we've funded, and ~they,~ they have ~a,~ a low cost braille reader, ~um, that,~ that they built. ~Um,~ and really trying to. Move the needle ~and,~ and ~bring,~ bring more opportunities for more people.

~Um, and it's super exciting stuff. Um, some of the ones that I'm most excited about are, um, Creating more opportunities for people with disabilities to engage in, in leisure activities. Um, which, you know, often when we talk about technology, it's, it's, there's a huge focus on productivity, productivity, productivity, or, you know, just like basic access to things like, you know, being able to find and get on a bus or navigate from place to place.~

~Um, and, you know, to, to actually be able to fund things that are like, how can we make. People with dis disabilities lives more enjoyable. And that's, that's super exciting to me as well. And even though the, the overall program is called AI for accessibility, not all of the projects are AI focused. Um, you know, a lot of it is innovation focused, although, you know, I, I do quite a bit of, of thinking in the, the space of AI as well.~

~That's awesome. Um, so I guess that's like a good segue so that~ I saw you have ~like~ an article, opportunities for AI and accessibility. ~Right. Um,~ so how are you thinking about AI in 2023 and ~like,~ like how's Microsoft approaching AI as well in 2023?

~I mean,~ I think any, anyone who has access to, ~uh, ~a,~ uh,~ newspaper or online magazine or what have you, can probably find out Microsoft's approach ~to,~ to ai, ~you know,~ AIing, all the things, ~uh,~ as we're recording this, ~um,~ Microsoft Inspire, ~uh, is,~ is going on and there's been lots of new announcements of new co-pilots and stuff like that.

~I think, you know,~ a lot of [00:10:00] the de development community is probably familiar with GitHub copilot and, ~you know,~ maybe somewhat familiar with more of ~like~ the large language model stuff like chat, G P T. And probably the image generation stuff, whether that's, ~you know,~ stable diffusion, whether that's Dolly ~or,~ or some of the others.

~Um, I think there's,~ there's huge opportunities in a bunch of different ways. So~ I,~ I will say upfront, I am fairly skeptical about. Ai. I think that there's a lot more that we have to do. ~I,~ I ~sort of, you know,~ I use the term AI loosely, ~you know,~ I, I don't believe that there is actually intelligence there.

~You know,~ some people can,~ you know,~ come at me on Twitter about that, whatever. ~Um, you know, I,~ I don't know that ~we're,~ we're heading towards the singularity, ~um,~ just because, ~you know,~ the tests that ~I've,~ I've done so far in, in playing with things just on my own time. I'm. Not overly impressed, ~you know, with,~ with sentence structure, for instance, in, in terms of ~like asking,~ asking chat j p t to write a couple of paragraphs about a particular topic.

~You know, I was,~ I was amused to find myself sighted when I asked it to tell me about progressive enhancement. But I saw it using a lot of [00:11:00] repetition. And when I asked it to talk about progressive enhancement in the context of various JavaScript frameworks, it gave me a rundown of various JavaScript frameworks, many of which did not ascribe to it progressive enhancement, but it made the same claims about their, ~um,~ their support for progressive enhancement, literally in the same sentence structure.

And it was just, it was such. Just cursory~ like,~ it, it reminded me of content farms.~ Like~ that's really what it felt like. And I'm not sure if that's because a lot of the content that was ingested to create some of those models was content farm generated content, ~um,~ which I see a lot ~in,~ in pitches for a list of part and such as well.

~So, you know, it's,~ it's something that ~I have,~ have a lot of familiarity with. ~Um, So I've,~ I've got a, a big skeptical hat on. That said, I do think that there is a lot of opportunity for these tools to be used in concert with our own processes. ~Sort of the, the,~ the human in the loop. ~Sort of~ concept. ~Um,~ and so I love the idea of co-pilots.

I love the idea of having, ~uh,~ tools that can make [00:12:00] suggestions for how we can improve things. ~Um,~ being able to, ~you know,~ ask, ~you know,~ GitHub copilot ~or,~ or some similar tool, like, how can I make this code more readable? ~Um, you know, are,~ are there potential. ~You know,~ issues with what I've written here that might create security vulnerabilities or something like that.

~Like~ that's super powerful. That makes us better programmers ~and,~ and honestly, ~it,~ it's a step. To the side and maybe a bit ahead of where a lot of developers are in terms of, ~you know, ~having a question, going to Stack overflow, getting the answer and just dropping it in without really thinking about it.

But they can actually ask the copilot, can you explain the solution that you have? ~Like,~ can you write comments to explain the code that you've just provided me with? ~Um,~ really having the tools there as, ~um, I've, I've,~ I've heard. LLMs described as a great improv partner, and so ~I,~ I like that idea of being able to, ~um,~ basically pair a program with, ~you know, a,~ a tool that has knowledge of a lot of different code bases and approaches to solving a given problem.

~Um,~ I think some of the coding results are really good. I think a lot [00:13:00] of the coding results are inaccessible and~ that's, ~that's somewhat problematic. And I think the, ~um,~ the proliferation of tools like this creates a lot of opportunities to, ~um, ~make things a lot worse in terms of really exponentially, ~um, you know,~ whether that's, Exponentially increasing the amount of inaccessible code or whether that's, ~um,~ expanding the use of ableist terms.

~You know, there's,~ there's lots ~of,~ of potential downsides to this. ~Um,~ but I think there's also opportunities. ~So, you know,~ in the, in that article I talked a lot about, ~um,~ Alt text because that's something that's been brought up as, ~you know, we've,~ we've had a couple of generations of image describers at this point, ~um,~ of varying quality.

~And, um, I think those,~ those will continue to improve. ~Um,~ but I like the idea of how,~ um,~ having that tool in the loop where it can take a stab at a providing alternative text for an image and. Even if seeing that prompts you to say, oh, that's not right at all. That's not where I, ~you know, want,~ want the focus to be in terms of describing this particular image.

Even if that prompt is just [00:14:00] like ~a,~ a nag at you ~to,~ to change it to something that you actually want, I feel like that's a net positive. ~Um,~ So I think that's a good thing. ~I,~ I could see a future where we, ~and I think I mentioned this in the article as well, where we~ could begin to enable people to interrogate images, interrogate graphics, things like visualizations, things like charts, ~uh,~ to be able to ask questions about it, to be able to, ~um,~ Even as ~a,~ a sighted person, for instance, be able to,~ um,~ simplify or change the format of a visualization, maybe I need to change it to accommodate the particular kind of colorblindness that I have.

~Um,~ or I want to reduce the visual noise in it by dropping out a couple of the lines if it's, ~you know,~ a line chart or something like that. ~Um,~ I think there's a lot of opportunity there. And then of course ~there's,~ there's huge opportunities in things like. I look at Volley, I don't know how familiar you are with that, but that's where you can, it's called, ~uh,~ few Shot Training where you can actually train a, ~um,~ an AI driven voice from just a few seconds of recordings of your [00:15:00] voice.

That is, Incredibly scary from a deep fake, ~uh,~ standpoint, but at the same time is potentially transformative for people who are losing use of their voice, whether they have something like ALS or something like that. ~Um,~ so I, I see that as being, ~um, you know, sort of~ a double-edged sword, right? And we need to be, it just, it, it makes it important for us to red team or catastrophize, ~however you,~ however you wanna look at it.

We need to think about ~like,~ what are the. Things that could be done with this that would be bad. ~Um,~ and how can we put mitigations in place to, ~um,~ avoid those risks, right? In order to be able to provide the most benefit with the least risk for people overall. ~Um,~ so I think things like that are just amazing.

~And then, you know, further on that, imagine somebody who is. Maybe nonverbal autistic that uses an AAC device to communicate. Um, maybe they can take, you know, sort of a, a stock voice. And because it's an AI driven stock voice, they can begin to manipulate what that voice sounds like in order to create, you know, a voice avatar for them.~

~Um, Which would also be really cool, you know, in, in places like Xbox and stuff like that, we can, uh, we can dress our, our avatars in different ways and give them different hairstyles and stuff like that. But what if we could do the same thing to, um, the ways our computers communicate for us, um, could be really amazing as well.~

Yeah, that's awesome. ~Um, quick time check. Kate, are we good on time? I don't, I don't cuz it cut off halfway, so I don't know how much time we did on the first half. Do you wanna just do a quick, quick talk about your book, Aaron, and then wrap up from there?~

~Yeah, I mean, I can, I can mention it. I mean, the book, the book's from 2016 now, so it's a, it's a little old, but, you know, it's still still useful. I still get people who are like, I just read it and it, it resonates as much to me~

~Yeah, I think, I think we should, we should just briefly touch on it and then we'll do the quick outro. Cool. Let's do it. Okay. All right. Awesome.~ So can you briefly tell us about your book, adaptive Web Design, and what exactly does a adaptive web design mean in 2023? Now?

~Yeah. So, um,~ I originally wrote Adaptive Web Design, the first edition back in 2010, and it came out in 2011. The second [00:16:00] edition I wrote in 2015, and it come, came out in 2016. So right around the time that progressive web apps ~were,~ were starting to be a thing. ~Um,~ and adaptive web design was ~sort of~ my way of.

Trying to encapsulate what Progressive Enhancement is all about. ~And, um, the, the subtitle of the book is Crafting Rich Experiences with Progressive Enhancement. And~ I feel like progressive enhancement is a bit of a mouthful. ~Um, so, you know,~ adaptive web design was ~sort of a,~ a good way to encapsulate ~that,~ that concept,~ um,~ and really put it ~in,~ in the framework of,~ um, You know, the, the, um,~ the Dow of web design and, ~you know, ~j John ops work and responsive web design, which when I was writing the first edition, ~uh, ~Ethan Mart's article, responsive web design, had just come out.

~Um,~ and so I was ~sort of~ thinking about it in that broad. Perspective, I guess as like ~here,~ here is ~a,~ a rubric for understanding how to build, ~uh,~ a web that actually adapts to the needs of ~your,~ your users. And in each of those additions, I really wanted it to be a philosophy book that. Had code examples to help you along with understanding what the [00:17:00] philosophy was and how that showed up in code.

~Um,~ and to not have it be a like, here's the latest and greatest way to do X, y, or Z. ~Um,~ I really wanted it to be framework agnostic. I wanted it to basically be able to have shelf life, which a lot of tech books frankly don't have. ~Um, and I've,~ I've been really impressed, ~you know,~ now we're what, seven years after the second edition came out?

And I still get messages from people who are like, I just finished reading Adaptive Web Design and ~you know,~ it spoke to me. It really helped me to. Put web design in perspective ~in a,~ in a different way. And, ~you know,~ it wasn't how I was taught in school or in the bootcamp that I did or what have you. ~And,~ and it really made me think differently about the way that I'm approaching what it is that I'm creating for the web.

~And I,~ I think that it makes me so happy ~to,~ to hear that,~ um,~ and to see that it's continuing to resonate with people because I, ~you know,~ I look back to. Books that really had a profound effect on me, like the pragmatic programmer ~and,~ and stuff like that, that really were more philosophy books. ~Um,~ and that's what I wanted to create.

~And so I've been, I've been really happy that the book continues to be, um, a source of inspiration for web designers and developers today. Um, so yeah, that's what's going on with adaptive web design. I've been asked if I'm gonna work on a, a third edition. Um, I'm not sure yet. I've got a seven year old, you know, he was, he was, uh, he was not born when I was working on the second edition.~

~Um, so, you know, it's a, it's a, a whole thing, but, um, but yeah, I'm, I'm sure it will be something that I, you know, here I am. Gosh, I think my first talk, or first my first article that talked about progressive enhancement was in 2003. So here 20 years later, I'm still talking about it. Um, I will likely continue talking about this approach because it continues to bear fruit, right?~

~It continues to be a valuable way to approach, um, design for the web in order to accommodate accessibility and to truly build inclusively, um, for the web. So it's, uh, it's something that continues to resonate for me and, and for others, which is great.~

~Awesome. Yeah, these topics are super important. So hopefully when your kid's off to college, you get the inspiration to, uh, to whip up the next one. That would be awesome. So, Look that about covers all of it. All of it. It was awesome talking to you. Erin, is there anything you want to shout out last, or where can people find you online?~

~Um, nothing particular to shout out, but people can find me on. Gosh. So social media is so splintered right now. Like I, I still exist on Twitter. I've been there since the very, very, very, very early days. Um, but I'm also on mastodon on front end social. Um, I'm am on. Blue sky, and I'm not on threads yet because I just can't, can't be bothered yet.~

~I'm, I'm, I've removed all social media basically from my home screen on my phone, so I, my, my consumption is a lot less. But folks can also find me on LinkedIn, um, and of course on my own web website and aaron hyphen gustafson.com.~

Awesome. It was awesome [00:18:00] talking to you, and thank you for

Yeah. Thanks so much, Chris.

~Yep. We made it.~

~How's, uh,~