From Sourdough to Strategy: Learning AI Through Passion

Tim O’Connor:

Hi there, and welcome to Kindrel Commons, where we take the mystery out of AI by showing you real people from around the world and how they are using it to make their life easier, smarter, and a little more fun. I’m Tim O’Connor, and sitting virtually right next to me is my co-host. He’s a man that is skilled in taking these complex systems of AI and explaining them in common sense.

That’s Dr. Adam Pah. Welcome, Adam.

Adam Pah:

Hey, Tim. Glad to be here talking with you, and thanks for the introduction. I like to think I focus on making things simple.

Tim O’Connor:

Well, you really are, Adam, and you’ve helped me learn a heck of a lot about these complex systems of AI. And today, we’re going to cover one of my favorite topics in the Kindrel Commons array, and that’s about how you can learn something new or get better at something you’re already familiar with with a little AI sidekick. And the guy that’s going to help us knows a thing or two about sidekicks because he’s been my boss formerly at three other companies.

He’s one heck of a business executive, but what you’re going to learn about him today, he’s also a world-class baker. We’re really glad to welcome you here to Kindrel Commons, Tom Pitera.

Tom Pitera:

Thanks, Tim. It’s great to be here.

Tim O’Connor:

Well, this is an exciting topic, if you will, this notion of learning something new or getting better at something you already are familiar with. So I wonder, Adam, could you talk a little bit about why this is so important right now?

Adam Pah:

It’s really unbelievably important because we’re really at this inflection point, right, where AI is becoming more and more usable. It’s becoming more and more dominant in what people, corporations, the world is talking about. But the reality is, is this question of, well, what is a generative AI assistant like Chad Chippity good for?

And really what it’s good for is helping us learn, right? It’s about learning new skills and having someone to actually kind of talk with, play with, and give us a greater worldview, help us answer the whys and dig deeper.

Tim O’Connor:

So this is more than just watching like this video or read this thread, but it’s something more, it’s a different approach, if you will, right?

Adam Pah:

Yeah. I mean, I don’t know about you, but I always describe myself as one of the worst students possible just because I do not learn by osmosis of being talked at. Like that just doesn’t do it for me.

But having an actual conversation and getting to ask questions and get answers back, that actually helps me understand. And it helps me actually kind of follow my own learning journey. And that’s what’s different here is we get to have our own personalized journeys where we’re getting this kind of feedback and this input that’s really specific to us and what we need.

Tim O’Connor:

So that feedback, that’s really interesting, access to feedback versus just access to an answer. Could you tell us a little bit more about that?

Adam Pah:

I mean, it really comes down to, I always think about being like a little kid, right? We always joke about that interaction, or at least it’s always on TVs, right? It’s like the little kid’s like, what color is the sky?

And the adult’s like, well, it’s blue. And it’s like, well, why is it blue? It’s like, well, I don’t know.

I don’t know, right? It’s the why, right? It’s figuring out and working through the why.

And we all have different whys. You know, there’s a million whys that happen anytime you get a fact. But it’s that follow-up and that greater understanding that actually transitions it from the rote memorization to actual, the journey towards comprehension, right?

That lets you actually take this and turn it into, well, now I feel like I can try it, as opposed to just also reiterating a fact and saying the sky is blue because the sky is blue, right? The fact is fact.

Tim O’Connor:

Well, I think what we can do is tee off of that to our friend, Tom Pitera, because everybody, Tom is absolutely an exceptional executive. He has run several multi-billion dollar businesses. He’s chairman of the board of two companies.

He’s worked as a board member, investor, operator, as I mentioned. You could say he’s made a lot of dough. Well, we’re going to talk about dough today, but a different type of dough.

We’re going to talk about sourdough. So Tom, you’ve been a long-term baker. Tell us just a little bit about your interest in baking.

Let’s kind of get that framework before like AI, if you will.

Tom Pitera:

Sure. It’s kind of been a hobby, right? For years now, I’ve started out cooking and I like to cook.

I’ve come from a big Italian family and my grandmother and my mother pretty much taught me how to cook. I’ve owned a couple of restaurants and I’ve learned how to cook and that kind of migrated into then learning how to bake, which is very different than learning how to cook because there’s pretty much a recipe that you have to follow and a process that you have to follow to get a good outcome in baking. And I think generalized cooking, particularly in Italian cooking, it’s much more freeform in the way you can go about it.

And so I started baking probably now eight, nine years ago. I know a lot of people got into it during the pandemic. I was baking long before that, actually probably even longer than 10 years ago.

And so I’ve been doing it for a long time. I enjoy doing it. And now that I’m working less and kind of semi-retired, I actually bake a lot during the day as well as I could spend a lot of my time actually doing that and have been doing that recently.

And so, yeah, I’ve been doing it for a while now.

Tim O’Connor:

You have tons of books already, right? Baking books and cooking books and stuff like that. What’s the problem with those types of things, right?

Tom Pitera:

Yeah, to your point, I have some of the best bread baking books that you can actually go buy. I have a book from Tartan, which is a big San Francisco sourdough bakery. I have Forkish.

I have Baking with Julia. I’ve got The Italian Baker. I’ve got The Breads of France.

And I probably got a half a dozen more books all about baking bread. But the problem with those books is that you end up reading the book. You try to follow the recipe, but there’s no feedback.

When things don’t go wrong or when things go wrong, I should say, or don’t go right, you have really no way to figure out what happened in that process. And that for me was very frustrating. And like I said, baking is a little bit of chemistry.

In fact, it’s probably a lot of chemistry. And when I thought the bread was going to rise really well and have a really nice crumb or a really good crust, and it didn’t, I couldn’t go back to the book and try to figure out why. And that’s where AI helped, actually.

Adam Pah:

I love the call out to chemistry there, right? Because they’re both chemistry. But you mentioned like cooking is more free form, right?

It’s like, you can add to it slowly, you can taste. But with baking, you’ve got one shot. You’ve got, this is it, we’re going and you’ll see in two hours.

Tom Pitera:

More like six or eight hours. Yeah, right.

Adam Pah:

It takes forever. So it’s, in a way, it’s your hobby, but it’s high stakes. And so how did you even feel about using GPT in the beginning here?

I mean, how did that even happen?

Tom Pitera:

So Gail and I, my wife Gail and I, we have four kids. And all four of our kids, in particular, a couple of them are really heavily embedded in using all of these large language modules. And I was talking to Tim, actually, one day, just having a conversation.

And I mentioned the fact that Matt and Emily both use these tools. And I know Tim was involved heavily in using these tools. And I said how, I was a bit reluctant, candidly.

And so Tim said, I’m going to give you an assignment. He loves to give me assignments. He always managed up really well.

And he said, I want you to use an hour a day. And I just want you to go and play with, pick your favorite tool that you want to use. And I use Chat GPT, mostly because it was free.

And so I just started to play with it, because that was the assignment that I got from Mr. O’Connor. And the other thing that he really told me, and this was, I think, really important, he said, don’t start to do it by talking about it with about business. He said, pick something fun and talk to it about something fun.

And as it just turned out, I had had a couple of failed bakes for sourdough bread. And so I started to talk to Chat GPT about, hey, why is my bread not turning out the way I want? Why is the crumb very dense?

Why is the crust not as thick and crusty as I want it to be? Like if I went to the Tartan bakery in San Francisco, or when I go to Italy and spend some time in Italy, the breads there are wonderful, or in France, the breads are wonderful. Why isn’t my bread turning out that way?

And I started to have a dialogue with the Chat GPT around why my bread wasn’t working really well and not coming out the way I wanted to. And I ultimately, as it turns out, multiple iterations of conversations with the tool. And I finally ended up loading in, just copying, pasting in the recipe that I was using.

And I copied and pasted in pictures of the failed bread that I was using. And it came back with recommendations. But it wasn’t judging me.

That was the other interesting thing. I didn’t get judged the fact that, well, you’re a knucklehead. Why is your bread failing?

Because the hydration’s not long enough, or you’re not bulk fermenting long enough, or you’re not overnight refrigeration fermenting long enough, or whatever it was. It was like, well, have you tried this? Or have you thought about doing it this way?

Or maybe you need to raise your hydration to 78% or 80% because you’re running at 72%. So it was a really interesting interaction on something that I was very much interested in, but ultimately was learning along the way. And different to your question, Tim, about different with the books, it was like going to a world-class Italian or French baker, or somebody sitting in San Francisco at Tartin Bread, and being able to talk to them versus reading it in the book.

And then not being able to get any feedback when things don’t go the way you expect them to. So it was a great experience and taught me a lot.

Tim O’Connor:

That is just fantastic. I would like to actually add on this one as being an educator you are. There’s so much we’re going to unpack with what Tom said, but let’s cover the whole thing about conversation.

Because when I talk to people that have tried GPT or quad or something else, they say they try and they fail and they walk away. And it’s kind of like the difference between communicating with Google versus communicating with an LLM. So can you talk to us a little bit about that and the differences there?

Adam Pah:

I mean, there’s a lot of differences, right? Because you’re not trying to retrieve information so much as you’re trying to have a conversation to surface an answer. Right?

We don’t treat talking with people like we do Google search, right? Or at least we don’t expect it to work well if we treat people like Google search. We need to give them some context, which Tom is doing here.

It’s like, I’m trying to do this. I love that it’s the photos is really what turned it around. Like the actual context of this is what is not rising.

And you know that you need to have some kind of intent of where you’re trying to take the conversation of like, I want to improve this. And a lot of it is just being open to iterating through the conversation with it. I mean, the reality is, as we’re talking with anyone, it’s not going to be 100% correct.

You need to be willing to experiment, try, report back and keep going. As Tom was talking, it really triggered this memory for me of just New York bagels and why they’re different and can’t be made anywhere else. And it’s because of the water, right?

And it’s like, well, how would you know that? If you’ve never been to New York and no one’s ever told you, all they did was let you eat a New York bagel and say, try to replicate this. It’s the details that matter.

And it takes an actual back and forth to surface that. And it takes details and explaining why you care, right? As opposed to just saying, give me an answer now.

Tim O’Connor:

Now, Tom, when you were communicating with it, as Adam just talked about, were there times, though, it came back and you said, that’s wrong? Or like, that doesn’t make sense to me? Or can you anything like that?

Because sometimes that happens, because AI is not perfect.

Tom Pitera:

No, in fact, it’s interesting that you say that. Because when you bake sourdough bread, you start with a starter. And that starter ferments for a number of days.

And then you add more flour and more water, and in some cases, a little bit more yeast and some salt. And you really are trying to achieve a level of hydration. And hydration is the ratio between the amount of water and the amount of flour that you use in the recipe.

And at one point, it came back and it did a calculation, because I was running, as it figured out, and I had known, but I was running at about 71% or 72% hydration. And it said, hey, to get the kind of results that you’re looking for, you really have to be at 78% or 80% hydration. And it did some calculations, and it told me how to adjust.

And I tend to be, as Tim knows, I tend to be a little bit mathematical in my approach and data-driven in my approach. And I actually found out that it was actually calculating it wrong, because it was not doing the starter hydration, because that was a different hydration than the actual added ingredients hydration. And so I went back and I said, I don’t think you’re getting the hydration right.

And it thought it for a couple of seconds, because it doesn’t take long, but thought for a couple of seconds, and it came back and it said, oh, no, you’re right. I did calculate it wrong. And here’s the adjustments that you need to make, because the tool wasn’t taking into account the correct hydration for the starter.

And it readjusted, and it readjusted what I was adding in terms of flour and water for the actual bread dough. So yes, it does make mistakes, and you do have to be monitoring the things that it tells you. But it was also very quick to react and say, hey, you know what, you’re right.

Tim O’Connor:

You know, it sounds, it’s like, in that case, it’s not you with a computer, but it’s you with actually like another person kind of sort of next to you, because it may make mistakes. You know, Adam, you’ve talked about a phrase, and I call it a very modern kind of mastery, iteration plus interpretation. It’s not teaching per se, but it’s helping you frame your own learning path.

Can you add anything to that? Because it’s not like that’s what Tom was having here.

Adam Pah:

Yeah, I mean, I just love that you’re pointing out here. It is a lot like working with a person, right? I mean, you know, no one does anything perfectly.

And so a lot of it is just continuing through this iteration and interpreting what it gives back to you, but then continually updating that context. And I mean, challenging it too. I mean, I love how Tom pointed out, it’s like, no, no, no, I think we’re getting the calculation here wrong, right?

But like, it’s not perfect, but this is how you learn, right, where the holes are and what its abilities are. But there’s this difference of it’s bringing you all this information and presenting it in such a way that you actually can act on it and learn quicker and go on that journey and get there with better results, hopefully, like Tom’s bread, or just get there quicker. No way to rush the bread, though.

Tim O’Connor:

You can’t rush the rise. Tom told me that. Maybe we’ll come back to that a little bit more on this one, Adam, and then we’ll come back to Tom.

And that’s this notion of, you’ve told me about strong prompts need three things, context, which you just mentioned, intent, and openness. Can you kind of explain what those are?

Adam Pah:

Yeah, I mean, so your context is really just setting the stage, right? Everyone talks about prompting or just says the word prompting. But the context here is really about you telling the story.

It’s you actually sitting down and thinking about, well, what are all of the details that I think could be pertinent? So you’re actually giving Chat GPT something to work off of here, right? Your intent is asking it directly.

This is what I want you to do for me. So give it all the information and actually make certain that you’re clear about what I want you to deliver to me. And the openness part of it is just being really open to the results that come your way.

I think you have a lot to say about this openness, Tim, but as I always interpret it, it’s about following up where it takes you, giving whatever your understanding is back and trying to clarify it and just being open to continuing the conversation.

Tim O’Connor:

Now tell me this notion of openness. When we started this conversation, you and I, about a year ago, right? And it was like, okay, do something you have fun with, like bread baking, because I knew you’d like to do that.

But now in terms of openness, how has this helped you to kind of transition from just being a better baker to actually transitioning over to using it in the work life? Because you’re chairman of the equity-owned companies and you’re always looking at acquisitions and improving the company. Can you kind of explain how it’s helped you by learning it through something you love now has helped you with something that you might also love, but it’s a different world.

It’s a different schema, if you will.

Tom Pitera:

Yeah, I think you’re right. The experience in working with Chat GPT on improving the quality of the bread or the outcome of the bread that I was trying to make basically gave me, I would say, kind of the confidence to use the tool in other areas and in other ways. And so I was much more comfortable than starting to ask about developing, kind of, or not so much developing, but how do we improve the strategy to go to market strategy for one or both of the businesses that I work with now.

Took it to the next step and really started to do with the Chat GPT the competitive analysis, right? Who are the biggest competitors that are competing against these companies and how are they differentiating themselves? And how have they differentiated themselves with pricing or what are the different strategies they’re using to differentiate themselves?

And the process though was very similar. It was still about having a conversation with somebody who was more of an expert and obviously had more knowledge about those areas, those business areas than I did. And again, it was very iterative, right?

It would come back and it would provide me with information. And then I would go ask another set of questions. And it was just that process.

But again, going back to the recommendation that you made very early on, just go try it with something that you love to do. But it allowed me to also learn how to do it for the things that, candidly, I would say that I needed to do. And it worked very well, but the process was very much the same.

And I think the points that you guys were just making about this iterative process, right? And setting the stage as to what are you looking for for a solution and what do you want the tool to do for you was really important. But again, spending hours and hours and hours talking to it about sourdough bread gave me the same opportunity and the same knowledge to be comfortable to ask those same questions about business problems, business strategies for these businesses that I work with.

Adam Pah:

Imprising to you to have it actually translate over from something fun but theoretically low stakes like baking to something we normally would call pretty high stakes with money involved.

Tom Pitera:

Yeah, a lot of money. A few dollars. Yeah.

No, I think that actually never really kind of crossed my mind. I think the reality was it was I had gotten, again, I had used it for, I mean, not excessively, but for a number of hours and then I’ve gotten comfortable with the process. So it was just, to Tim’s point, it was just migrating the process to something that was just different, different topic.

And knowing, candidly, you have to verify, but knowing that there was the information that was coming had to be verified and it had to kind of fit, did it pass the sniff test, right? What is it telling me? Does that really make sense to me, given what I know of the business and the industries and stuff that we were talking about?

So it was a really seamless and pretty easy transition to be really candid.

Tim O’Connor:

Let me ask you about something about regionality. So I’m going to ask you first, Adam, and then we’re going to transition back to Tom about you can’t rush the rise, okay? So Adam, one of the things that happens is that Tom was giving all these questions and answered conversation, et cetera.

Sometimes there’s an effective notion of regionality, right? Like how does it apply to me here in this place, in location, time, space, whatever. Can you add a little bit about that when you’re doing prompting to make sure that you can get your answers more fine-tuned, more precise to who you are, what you’re doing?

And then I’ll come back to you, Tom, about can’t rush the rise.

Adam Pah:

Yeah. So I think a lot of that goes to actually being really thoughtful and specific about providing that context. That’s what I was kind of thinking about when I mentioned the New York bagels and the water, right?

There’s a difference depending on if Tom was in Florida, right at sea level on the beach trying to make bread, or if he’s at home in the Midwest and it’s the heart of winter, right? All of those details add up a lot. And honestly, before these tools, that’s what we were missing from tools is the ability to get this kind of personalization to our exact circumstances in this specific moment, in this specific time.

But the reality is you have to give those details as a part of your context, or at least ask questions about if they could affect things. If you’re going to get that kind of personalization that you’ll get the improvements from. And that’s really the difference here is we’re talking about moving away from getting the average answer that’s correct for the average person in whatever the average baking place of the world is, right?

And we’re instead talking about for Tom that’s in Florida right now or in Ohio right now, what’s the best course of action here?

Tim O’Connor:

Now, Tom, when you’ve done interactions with GPT, have you spoken about location?

Tom Pitera:

Yeah, you have to, because to your point, you can’t rush the rise. And it’s interesting, I can be sitting here in my house in the middle of the summer, and the environment is very different than sitting in the same kitchen here in the middle of the winter. And there’s some things that you can correct for and adapt to.

But I mentioned this Tartine Bakery in San Francisco, and there’s this, they keep their windows open at night, and they allow their starters to get that damp, moist Bay Area air into their bakery, and it impacts their starters. Well, I can’t create that here, right? I don’t have that Bay Area atmosphere, but I can adjust other things like, you know, in the summertime, the ambient temperature in the house and the humidity in the house is a lot different than it is in the wintertime.

And so I have to adjust for that. And I feed that information, or I discuss that information with GPT, and said, hey, you know what, I’m sitting here right now, and it’s 84 degrees outside, and the relative humidity is 62%. And that’s very different than if I was sitting here in the winter, and it was minus five outside, and the humidity was at 5%.

And you get back, like you were talking to that professional baker who knows how to adjust for those conditions, you get back that information, because you fed it those specific thoughts, right? And those specific things that are happening in the environment right at that moment. And that’s no different than if I was to transition it to the business side, right?

Because business is dynamic as well. So what’s happening in the world today in this industry or for this business is going to be very different than what happened six months ago or six months from now. And so I think being very specific and prescriptive is really, really helpful when using these type of tools.

Tim O’Connor:

And before there was GPT, Tom was GPT for me, because I went to Tartine Bakery, I ordered some bread for Tom, because he had never had it from there. This is many years ago. And you order it a couple of days before, and then you show up at a certain time, like precisely at 11 o’clock or 12, whatever it is, you show up.

And if you’re not there within like 20-30 minutes of when you’re supposed to be there, they’ll sell the bread to somebody else. I get there. And there’s a sign, the bread’s not going to be ready for like another 30 minutes.

And I’m like sweating because I have to take the bread over to the FedEx at the airport because I’m going to go ahead and send the bread to Tom. And I’m like wondering what’s going on. Anyhow, I got it to Tom, et cetera.

But I was asking him, what’s going on? And Tom told me, like GPT would have said, you can’t rush the rise. And he explained it all just like GPT, because I had no clue what was going on.

So before we do our closing thing about Real Talk, now try this and teach one more thing. Tom, anything else you want to add? And then the same thing to you, Adam.

Anything, Tom, to add before we do our closing?

Tom Pitera:

I think, Tim, I would go back and anybody who is hesitant on using these type of tools, the advice that you gave me a year, year and a half ago would be sit down, go play with the tools, have a conversation, have a conversation around something fun that you want to learn more about or that you just want to converse about. And I think if you do that for an hour a day for the next two weeks, you’re going to find out that you’re going to have an amazing experience that’s going to, I mean, candidly in some ways, some small ways, but change the way you do things. Honestly, I don’t Google search anymore.

No matter what I’m trying to find, I use GPT. And I think that transition for me was really, really amazing because I used to use Google all the time. And I think one of the things that companies are going to have to figure out here is that’s a pretty big dynamic shift, right?

Because one of the businesses that I actually do business with is heavily reliant on Google search. And that’s being impacted pretty dramatically from a business perspective because of these language modules and models. And again, I think we could have another whole discussion about what the impact is and how do you ultimately monetize some of these new tools that are in the market like GPT, but go back to what I said.

I think the advice that you gave me early on was to just go have a conversation with it. And candidly, if you do that for an hour a day for the next two weeks, you’re going to find that you’re going to have an amazing experience. I’m pretty confident of that.

Tim O’Connor:

Well, that’s super. Well, actually that does a great job at answering the question of real talk, like what would you recommend to somebody? So that’s super, Tom.

Thank you. Adam, any other comments before we get to the other two things, which is next thing to do and tell somebody else?

Adam Pah:

I think I just want to really glom on to what Tom said there, but really to highlight the fact of starting with something that’s really low pressure is kind of key when you start these journeys and these conversations, right? Start with something that you really care about. Start with something that you want to experiment with or you’re willing to actually follow up and get that kind of feedback and improve, because that’s how you actually learn what the tool is good for.

And that’s how you actually take this journey. And it’s important, right? It’s everyone’s job to always grow, become better.

At least most of us want to become better, I think.

Tim O’Connor:

Well, that’s a really good answer to our second thing in the closing, which is now try this. So basically what I’m hearing is like pick something. Pick it now.

Just go in and try it. It could be cooking. It could be pickleball.

It could be organizing your garage. Doesn’t matter. Something that might mean something to you, but just try it and put that question into GPT and start having a conversation.

Is that kind of the thought?

Adam Pah:

Yeah, I think so. I mean, it’s really just this thing. If you care about it, learning is contagious, right?

You’ll want to keep using it. You’ll keep growing with it. And once you build that muscle, it really makes it easy to transition from one thing to another.

Tim O’Connor:

Well, and our last area where we kind of close is all about teaching one more, right? The way you build these things is teach somebody else what you’ve learned. I know, Tom, you’ve had some conversations with one of your sons about it, though.

He’s taught you some things. Have you taught him anything along the way?

Tom Pitera:

Yeah, he’s taught me a lot more than I’ve taught him in this area. But I’ll give you one example of teaching somebody else. And that was my wife, Gail.

And one of the things we were doing and have been doing is planning a family vacation to do a safari. And Gail spent literally hours and hours and hours over a course of multiple weeks getting information about putting together a safari. And she was showing me some of the information.

And I said, this was all really good stuff, a lot of really good work, a lot of really hard work and dedicated work to try to make this family vacation come together. And then I sat down in front of ChatGPT, and I said, here are the things that I want. Here’s what I’d like to know.

I want to go see this. I want to see the big migration. I want to see the big five.

When should I go? Who should the outfitters be? And literally, I walked out in 15 minutes later and put a printout in front of Gail that took her literally 20 or 25 hours.

And she said, how did you do that? And I literally, we sat down and I showed her how to use ChatGPT the way I had learned it by that point in time or enough of it by that point in time. And I’m not an expert by any means.

But, and it was amazing. Now she uses it because she figured out how it really made her life in a lot of areas so much better as well. So I think that, I think your point, Tim, about showing somebody else the benefits is really beneficial, not only for that individual, but also for you in terms of how do you tell somebody else about it and teach somebody else how to use it was pretty important, especially in this case, because it happened to be my wife.

And she was pretty upset when I did it in 15 minutes, but was pretty happy when she started to figure out how to do it herself.

Tim O’Connor:

That is an absolutely wonderful way to end here. You know, and our basically recipe, if we stay with bakery, is real talk, try something, teach one more. So what would you tell somebody else?

Do something right now and teach one more. I mean, that is a recipe how this grows. That’s how we build Kindred Commons, one conversation at a time.

Tom, it’s been an absolute pleasure having you. Thank you so much for joining us today.

Tom Pitera:

It’s been my pleasure, Tim. Thank you. Thanks for Adam too.

Tim O’Connor:

And Adam, always my partner in this. It’s great having you. Thanks so much for all the insights for everybody.

Adam Pah:

Well, thank you, my partner in crime. Thank you, Tom, for sharing so much with us. It’s been a great talk.

Tim O’Connor:

And thank you everybody for joining us here at Kindrel Commons, where we believe AI is for everyone, not just a few. And we’re learning something new might just start you on your next loaf of bread or something else. So see you next time.

And thanks for joining. Bye-bye.