Technology develops differently in Japan.

While US tech giants have been grabbing artificial intelligence headlines, a business AI sector has been quietly maturing in Japan, and it is now making inroads into America.

Today we sit down again with Miku Hirano, CEO of Cinnamon, and we talk about how exactly this happened.

Interestingly, Cinnamon did not start out as an AI company. In fact, when Miku first came on the show, the company had just launched an innovative video-sharing service. Today, we talk about what lead to the pivot to AI and why even a great idea and a great team is no guarantee of success.

We also talk about some of the changing attitudes towards startups and women in Japan, the kinds of business practices AI will never change, and Miku give some practical advice for startups going into foreign markets.

It’s a great discussion, and I think you will really enjoy it.

Show Notes

  • How Miku invented TikTok before TickTok and why it didn’t work
  • How you know when  its time to pivot a startup
  • Why companies will never go digital and will always use paper
  • Who will benefit most from AI
  • The four categories of AI
  • How AI will change the legal profession
  • How japan is actually ahead of US and China in some kinds of AI
  • What’s really driving business innovation in Japan
  • Can AI actually reduce overtime?
  • How enterprise clients treat women founders

Links from the Founder

Leave a comment


 Welcome to Disrupting Japan, straight talk from Japan’s most successful entrepreneurs.

I’m Tim Romero and thanks for joining me.

Today, we’re going to sit down and talk about artificial intelligence with Miku Hirano of Cinnamon. Now, Cinnamon is actually a great example of a successful Japanese startup pivot. When we first sat down with Miku four years ago, she had an innovative micro-video sharing company called Tuya and really, you should go back and listen to that episode. I’ve put a link on the show notes and it was really a good one.

Anyway, Miku basically started TikTok a few years before TikTok and we talk about why things didn’t work out, why even with the same idea, one startup will become a multi-billion dollar brand and the other will pivot. Of course, the pivot to AI and the rebranding to Cinnamon has led this to their current success in using AI to read and to understand common business forms.

In fact, for reasons that Miku will explain during the interview, Japan is actually ahead of the US and China in the area of business AI. We’ll also talk about how attitudes towards women are changing here and how Japanese men at traditional companies treat women founders, particularly women founders with children, and I think it might surprise you. I mean, it surprised me and it surprised Miku as well,

But you know, Miku tells that story much better than I can, so let’s get right to the interview.


Tim: So, I’m sitting here with Miku Hirano of Cinnamon and it’s great to have you back on the show again.

Miku: Yeah, thank you so much for having me here again.

Tim: Well, so much has changed since — it was three years ago, right?

Miku: Yeah, yeah, and I had a totally different business at the time.

Tim: Well, not only a totally different business but you’ve gotten married and you’ve had two kids.

Miku: Yeah, yeah, and at the time, I think I was living in Taiwan and now, my business is in Tokyo, so everything has changed.

Tim: And so, we’re not even going to cove what we talked about last time even though in the intro, I’ve told my listeners to go back and listen to the old show, I know some of them won’t do it.

Miku: So, please listen to the last audio show.

Tim: Okay, so guys, listen to Miku. It was a great show.

But let’s just pick up where we left off because it’s been a busy three years. Last time we talked, you were working on Tuya which was this great seven-second video sharing startup. So, what happened?

Miku: Tuya was a video sharing app but unfortunately, that business didn’t go well, so we couldn’t get enough users to raise a fund, so I called the Taiwan office at the time and…

Tim: Let’s get back to it because I mean, the mechanics of what happened are important, but I think what’s more interesting is the idea, the seven-second videos, it was a great idea. It was basically TikTok a couple of years before TikTok, and this is something that happens so often, so for example, like Uber and Lyft were not the first companies to try that business model. AirBnb certainly wasn’t the first company to try that business model. There is so much more involved than just having a great idea and a great team, and we’ll get to how you’re executing now, so we know it’s a good team.

Miku: Thank you.

Tim: But why do you think that Tuya didn’t take off where TikTok became a worldwide craze?

Miku: I see, so we lost some users and some of the users used Tuya, our app, every day. They took a video, it’s just a seven-second video maybe seven times or eight times a day, so there are some active users, but I think Tuya doesn’t have a power that our users want to tell to other friends. So, for example, TikTok, sometimes, you find some very interesting hilarious videos, right? And sometimes, you show it to your friends, but on my app, we didn’t have that part.

Tim: So, was the key innovation, the key driver at TikTok, you think, was that integrated sharing mechanism?

Miku: Yes, yeah, to other friends.

Tim: Alright, and what was the trigger? What made you decide to say, “Okay, we’re shutting down Tuya and we’re moving to AI”?

Miku: So, that pivot didn’t go that quickly. It was very ugly, to be honest.

Tim: It usually is. So, what were the steps? How did it roll out?

Miku: Okay, so our company was almost going bankrupt, so we found that our company will have no cash in two or three months, so we needed to change our business.

Tim: And you were a pretty big company at that point, right? How many employees did you have?

Miku: We did have around 20, but I need to restructure maybe one-third of the company, and then I need to come back to Tokyo to survive some of the members, I need to earn money, so I came back to Tokyo. So, in that point, I had two options. One is to raise enough fund and the other is to get cash, I mean, from sales, but for the first option, I guess, I thought it wouldn’t go well. We just had a failed business, so both me and Hajime, my co-founder Hajime are serial entrepreneurs but we couldn’t make the second business wait.

Tim: Yeah, I think especially in a business model like yours, raising funds at that point would be extremely difficult because the investors would say, “Well, look, you tried it, it didn’t work, try something else.” Did you try to raise funds or did you just –?

Miku: Actually, I didn’t try.

Tim: Just figured that this is time to…

Miku: Yeah, just finding, I mean, just doing sales would be better.

Tim: What kind of sales did you do?

Miku: Just system development. So, at that first point, we didn’t think about AI at all.

Tim: So, just we’ll develop apps, we’ll do javascript programming app development business?

Miku: Yeah, with somebody else to get money. I think most of entrepreneurs having engineering background have this kind of idea, that if our business doesn’t go well, we can do systems development business and then we can survive, so we have that kind of fault, but actually, it didn’t go well.

Tim: It did not?

Miku: No, no.

Tim: What happened?

Miku: It was really difficult. Yeah, so I did not have any sales experience and also, another reason, most companies have experience, have system development company that they have already used, right? So…

Tim: Right, so they need a reason to start looking for a new vendor.

Miku: Yeah, yeah, so yeah, I improved the sales material that there was a page, things that we can make, and actually, one of them was AI. The reaction of the potential clients got changed.

Tim: You found something that their current vendors couldn’t do.

Miku: Yeah, and at that point, we totally forgot that both me and Hajime didn’t research about AI when we were students. We totally forgot about it.

Tim: You did your Masters in AI, didn’t you?

Miku: Yeah, yeah, and even my co-founder had phD in AI deep learning, but we totally forgot about it.

Tim: Just too much stress of trying to keep the company going?

Miku: Yeah, and then to increase what we can do, I just wrote AI, then their reaction changed, so I thought it might be a chance.

Tim: So, you started getting contracts, and then you just started re-tooling the company around AI?

Miku: I started to get contracts from a couple of companies, so that is how we could survive.

Tim: And I guess from that interest, eventually developed the Flax Scanner product.

Miku: Exactly, so the idea of Flax Scanner which is our flagship product came from one of our projects. So, one of the projects was extracting the information from resume.

Tim: And so, who’s using it now? What kind of companies?

Miku: Yeah, so now, more than 50 companies use Flax Scanner, mostly insurance companies and banks, and manufacturers, including Toyota, Daiichi Insurance, Nihon Insurance, so that kind of top tier Japanese huge enterprise.

Tim: And what kind of documents are they using this on?

Miku: For example, in an insurance company, the receive tons of documents, so for example, the identification cards or you need to send that document when you get into a hospital or when you have a car accident, so they have millions and millions of paper documents.

Tim: Certainly, using AI to read in paper forms is really valuable, particularly with the amount of data that’s already on paper, but do you think this is kind of a transition step or I mean, so for example, do you think eventually, everything will just be digital to digital or do you think that we’re going to be stuck with paper forms for a long, long time?

Miku: Yeah, so as you know, most of the companies have tried to reduce papers for maybe the past 20 years.

Tim: Yeah, MEs, yeah.

Miku: But it doesn’t go well.

Tim: No.

Miku: And yeah, of course, some of them are getting digitized but yeah, we use some papers, still. So, for example, one of our clients, Daiichi Insurance, they have 2000 people just only for data entry work.

Tim: And that’s just reading paper forms or paper documents of different kinds and putting it into the computer?

Miku: Yeah, yeah.

Tim: Wow. So, is your basic strategy now to focus on Flax Scanner or to use Flax Scanner as kind of an entry point and to build vertical services within, say insurance or within banking?

Miku: Yeah, we did both, yeah. So, we found insurance companies and banks in that kind of finance industry. They have points that we should target on, so now we are rolling out kind of the same thing to other insurance companies, so for example, there are 200 insurance companies in Japan and we have only got purchase from 20 insurance companies. Even outside of Japan, for example in the US, they do the same thing, so we only have a sales branch in Austin, Texas. We found out that the industry structure is exactly the same as in Japan.


Tim: What I find fascinating about this is using AI to read paper documents, to read paper forms is — when I was first starting my career in technology 30 years ago, using AI to read paper documents was one of the most promising aspects of AI. Are reading forms fundamentally that difficult?

Miku: Yeah, so I guess people might thing reading papers is kind of easy version of AI, I guess? But actually, it is very difficult and I think there are two difficult points. One is reading letters, so actually, many companies have thought about using OCR. Actually, five years ago, but at that time, in OCR technology, they didn’t use deep learning, so they used only pattern matching, so if the font is different from normal thing, OCR doesn’t work well at the time.

Tim: Okay, so very small changes would throw it off.

Miku: No. Two years ago, people started to use deep learning for OCR, and then accuracy improved a lot.

Tim: Is the real challenge in this recognizing the character shapes or trying to figure out what it means in context? What’s the difficult part of this problem?

Miku: Yeah, so that is a second thing. On a paper, there is so many information. So, if you think about invoice, if we have two invoices from different companies, the formats are totally different, right? So, for AI, it is already very difficult to understand what is what. Usually, on an invoice, there are, I guess, 20 or 30 items, and each item is different, right? So, for example, this item means a company name, this item means company address, and this item is bank information, so to understand that, actually, natural language understanding is needed.

Tim: You know, it’s really interesting how AI is getting applied to solve — coming from an engineering point of view, this kind of information exchange should be very easy to standardize and to solve, but in reality, and many attempts have been tried, but in reality, you’re right — everyone formats things differently, everyone does things a tiny little bit differently that I guess without AI, you have to use people and that is expensive.

Miku: Yeah. So now, 2,000 people in one insurance company are working on just only for data entry work.

Tim: You mentioned before that something that seems really easy to people are actually incredibly hard for AI to do, so what kind of tasks is AI good at and what kind of things is it just bad at?

Miku: AI is good at repetitive tasks, yeah. I think people think about AI as all too much like magic, so people think you can do anything with AI.

Tim: What industries? What kind of companies do you think are going to benefit most from AI?

Miku: It is a very wide question because first of all, there are four AI categories. One is Internet AI and the second one is business AI, and the third one is recognition AI, and the last one automotive AI, and Internet AI is, for example, TikTok or Uber kind of apps. I mean, in those web service, they use AI.

Tim: So, it’s fundamentally optimization, recommendations?

Miku: Yeah. So, for example, TikTok, there are so many videos, right? But how to show videos to you?

Tim: Okay.

Miku: So, like that, and the second one is business AI and this is for supporting office workers and Cinnamon is in this category, and the third one is recognition AI, for example, finding strangers in — I mean —

Tim: Facial recognition type AI?

Miku: Ah, yeah, facial recognition or bio security camera. It finds strange people or finding sick from medical images, so that is recognition AI, and the last one, automotive AI is, for example, self-drive or AI speaker, those kinds of things, so depending on the category, it’s different.

Tim: To hear you explain it that way, I mean, when you think about it, there really isn’t such a thing as AI. It’s this big collection of different technologies and different approaches that are being used to solve very different problems. It’s just, we have a very trendy marketing word that gets to be applied to all of them.

So, Cinnamon is obviously focused on the business AI.

Miku: AI, yes.

Tim: So, let’s focus on that one. Who do you see benefiting most from business AI?

Miku: Yeah. So, in business AI, I think yeah, there are so much varieties, so I guess most of the companies can have that benefit. For example, one of our projects that we did with a bank, that client quite often invests in foreign companies and their investment contract has 400 pages or 500 pages and a lawyers spends, for example, one week to extract 100 points from that huge document, and then after that, the team discusses based on that 100 points, but our technology can do it within 20 minutes.

Tim: I think yeah, AI is going to revolutionize and change the legal profession just as much as spreadsheets changed accounting in the 80s and 90s because I think we’re almost at the point — maybe we are already there where AI can understand a contract better than the average attorney can because it is such a specialized stylized language.

Miku: Yes, yeah. For humans, lawyers and doctors, accountants, they are kind of high-skilled people, and AI in some parts AI can just do it.

Tim: Well, yeah, I think, and if you look at what happened with the accounting in the 80s and 90s, this whole army of bookkeepers just disappeared, and the only people that remained were the consultants and the people who could provide sort of financial strategy and that high value, and yeah, it would make sense if law, the legal profession, kind of went the same way.

Miku: Yeah, but there are some things that AI cannot do, so there are three things. One is management and the second one is creativity, and the third one is hospitality. So, for AI, actually, kabakurajyo is very difficult to replace.

Tim: I think so, yeah. I mean, in terms of creativity, since the 70s, there’s been computers that can compose music or paint, or something like that.

Miku: Ah, yeah. So, there is some AI to make music. So, AI is good at doing repetitive things, so for example, there are so many songs that Beethoven made. If we train AI to learn the past Beethoven songs, actually, that AI can make a song like Beethoven.

Tim: Yeah. I mean, AI has been really good at doing like, fugues, Bach style because it’s so rigorous.

Miku: Ah, I see, I see. Yeah, so now, AI can make, for example, not only Beethoven but also Chopin, like some kind of —

Tim: Styles to play.

Miku: How do you say? Not very rule-based.

Tim: Right, right, okay.

Miku: Yeah, but for AI, it is impossible to make new types of songs, so for example, if you go back to the 15th century in Japan and even though we bring AI to that point, if you do not have any songs that Beethoven made…

Tim: It won’t come up with it on its own?

Miku: We cannot.

Tim: Right, right. A lot of people say that Japan is kind of behind in AI, particularly compared to the US and China, and what are your thoughts on that?

Miku: Yeah, so to be honest, Japan is a lot behind compared to China and the US, so I told you there are four categories of AI, but actually, in business AI, Japan is a lot advanced.

Tim: Really?

Miku: Yeah. Actually, Japan is the most advanced country in the world.

Tim: Why is that?

Miku: I guess there are two reasons. One is Japan is doing so inefficient things now, so they need to be efficient.

Tim: Japan has the most paper — it has the most demand for reading forms. That actually — yeah, I could see that.

Miku: Yeah, so I have seen a ranking before about automation potential ranking and Japan is number one, but basically, I think that ranking means inefficient ranking.

Tim: Well, I don’t think it’s that different. I think it’s actually measuring the same thing because if you look at where anytime innovation takes off, it’s where there is the most need. So, I mean, working a lot in energy, for example, and all of the really interesting innovative energy companies are not in Japan or the US. They are in sub-Saharan Africa or Latin America where there is a really strong me, and I guess Japan just has a really strong need.

Miku: And the second reason is open innovation. Yeah. I feel two years ago, most big enterprises already have, for example, innovation department. They started to have a big mission to find good technology from outside of the company.

Tim: So, you think Japanese companies have got more open to working with startups recently?

Miku: A lot more than before. So, in my first startup, Naked Technology, actually, in the very beginning of Naked Technology, we tried to sell AI because me and Hajime, we had a background of researching AI, and of course, at the time, we tried to reach enterprise but it was impossible for us to set a meeting with a big company.

Tim: Why do you think it changed? What do you think made it changes so fast?

Miku: I don’t know. It wasn’t fast because more than 10 years have passed, right? So, is it fast?

Tim: I guess so. You know, it depends on what you’re looking at, so like my first company, I started here in 1998, and it was really, really hard to sell to big companies, so I guess from my point of view, it seems like the change happened very quickly.

Miku: Maybe in the past 10 years or 15 years, Japanese companies have tried to do cost down, and in these 10 or 15 years, we have done everything that we can about cost reduction.

Tim: And so now, they are looking at new technologies?

Miku: Yeah, yeah.

Tim: Alright. So, you mentioned you have an office now in Austin. Are most of your customers from Japan or do you have a pretty global customer base now?

Miku: Still, more than 90% of our revenue is coming from Japanese companies, but we found the exact same situation in the US and yeah, of course, we talked with our competitors as well, but we feel the US is one and half years behind compared to Japan.

Tim: In this business AI?

Miku: Yeah, yeah, in this business AI.

Tim: And it’s targeting the same kind of industries, insurance and banks?

Miku: Yeah, exactly.

Tim: Huh. Now, a lot of Japanese startup founders have told me that VCs tend to discourage startups from going global until after the IPO and after it’s safe. So, your recent investment round, a lot of that money was earmarked to go global. So, did you feel any of that pressure from the VCs or your investors very supportive of the idea of going to the US?

Miku: Ah, they are supportive. One of the reasons that I started this company Cinnamon is to try global because in my previous company, we wanted to go global but we couldn’t, so if we cannot try global expansion, it is not meaningful for me, so to all of the investors, I have already told that we want to do global expansion.

Tim: So, was it just the investors knew what they were getting into from the beginning or do you think there’s something about Cinnamon that made them more confident that it could work in the US?

Miku: Yeah, I think so, especially all of our management have experience in global management. So, I guess our investors are feeling that probably, the possibility is higher than other startups.

Tim: Okay.

Miku: Maybe.

Tim: Alright, and what made you focus on the US market rather than, say Southeast Asia? Because you have such strong ties to Singapore and Taiwan, and the US market entry is so expensive.

Miku: Yeah. It’s super expensive, but still, the US is the best market after Japan because US has three times bigger market in China has four times bigger market than Japan, but China is a lot behind in this business AI. Actually, we tried to find competitors in China but we couldn’t .

Tim: That’s surprising.

Miku: Yeah, so I guess sexy AI, like self-drive or AI speaker, those kinds of startups can get one hundred million dollars.

Tim: So everyone is focusing on the sexy AI and you’re focusing on business AI?

Miku: Yeah, so business AI is –

Tim: Wait a minute! You named your first company Naked Technology.

Miku: It was too sexy

Tim: That’s a pretty sexy name.

Miku: Yes, so I guess business AI, if you have a business AI company in China, I guess it is very difficult to get the money.

Tim: All of my companies have been business-to-business sales. I’ve always liked B2B. It’s simple. Consumers are crazy and unpredictable, but business, at least on average, tend to behave rationally.

Miku: Yeah, yeah. In the previous business was at business, right? And yeah, I felt it was very difficult to predict the users’ behaviors, but at least for business, they calculate how much they invest and how much they can cost down, or increase revenue.

Tim: Yeah, the sales, it’s simpler, I think, right?

Miku: Yeah, yeah.

Tim: So, how are you doing sales? So now, you had to go through — you had to get good at sales over the past three years. I noticed you’re doing a lot of kind of speaking events. Is that useful for driving business?

Miku: Yes, so 30% or 40% of our clients are coming from our homepage, so they write something on our form on our website.

Tim: Really? So, business-to-business but it’s mainly inbound web-based marketing?

Miku: Yeah, yes, and I think 20% to 30% are coming from exhibitions, so for example, AI Expo, we have a booth and then people come to our booth, and then we exchange business cards, then it became leads.

Tim: Okay. I don’t want to keep you too long but I want to talk a bit about you.

Miku: Okay, sure.

Tim: So, last time we talked, we talked a lot about women in startups and women in tech, and so much has changed and you have like, two babies now, and so I wanted to ask you how that’s changed things, how you have managed to run a startup with two little kids and what’s been the attitude of both business partners and potential customers? How do you make that work?

Miku: Yeah, first of all, it is super tough.

Tim: I imagine, yes.

Miku: Having a business and having a two-year-old and a one-year-old, it’s super, super tough. Now, I do not have any time for myself. So, working and dealing with my kids, and then sleeping, that’s it, so that’s my 24 hours, so it is tough, but having kids gave me so much influence. Our company mission is coming from both of my kids and at that point, I read a sad news about one Japanese young lady committing suicide because she worked too much, karoshi, and I guess if I read that article before childbirth, probably karma I would’ve thought yeah, it is sad, but I guess I would forget about it after that, but I started to think we shouldn’t have this kind of a working environment for the next generation. This working style should not remain to my kids.

Tim: So, you think AI is something that can sort of help?

Miku: Yeah. So now, everybody is busy every day, right? But if I think about myself, not all my working time is very productive. So, sometimes, I do very repetitive tasks which I don’t want to do, so I thought if AI could do that part, my working time could be short.

Tim: This is something I wonder about because I think as startup founders, you and I and most of the Disrupting Japan listeners are used to trying to optimize every minute of our time. We can never do everything we want to do , but I wonder how much, particularly in Japan, of the overtime work really is people who don’t have enough time to do their job and how much of it is just that it’s culture to do overtime and to find work to do?

Miku: In that sense, I think Japan is changing. I think it is because of Abe-san’s workstyle revolution and now, people start to think about working too much time is not a good thing, and so culturally, I feel it is changing.

Tim: I think a lot of that is generational too. I think young Japanese in their 20s and 30s won’t put up with that as much as the generation before what. Do you think Japan is being more accommodating and open to working moms, to people who are trying to balance children and career?

Miku: Yeah, yeah, I feel so, and also, Japan, we will not have enough people, enough labor in 10 years, so we need 10 million people more to maintain our current business, so we need to let mothers work and AI to work.

Tim: I mean, that’s, from a high-level, I think everyone agrees with that, but there’s always a gap between the high-level, it would be best for Japan if Japan becomes more open towards working women and working moms, and then the real everyday interaction with people, and do you think like the real everyday interaction with business partners and customers, you felt that that changed, that people are becoming more open to it?

Miku: I feel so. So, for example, when I was pregnant, our potential clients came to our office because they thought it was difficult for me, it was tough for me to…

Tim: Oh, that’s great!

Miku: Yeah, like that, and they are in a lot higher positions as me, and they are potential clients, so I was supposed to go to their office.

Tim: Yeah, on several different levels, yeah.

Miku: So, for example, they came to our office.

Tim: That’s great to hear. No, that’s exactly the kind of thing I was asking about. That is a real shift in attitudes. So, customers have been supportive as well. That’s fantastic.

Miku: Yeah.

Tim: What about your parents? So last time we talked, you were saying that your parents kind of secretly wanted you to be more of a traditional ojou-sama and…

Miku: They failed.

Tim: Yes, at last time we talked, they were still kind of disappointed with — have their opinions changed and now that Cinnamon has become successful or maybe now that they have grandchildren?

Miku: Yeah, they changed. So now, yeah, fortunately, some media covered me, so they sometimes see their daughter on the newspaper that they subscribe to, so now, they feel, yeah, they are proud of their daughter.

Tim: And they understand what you’re doing is a startup?

Miku: Yeah.

Tim: That’s fantastic. It seems like there’s been this really big shift in attitude. Again, to me, it seems really fast, something in the last five or 10 years, and back when you were at Todai, you were the only woman in your CS class, right?

Miku: Yes.

Tim: Is that changing?

Miku: I don’t know, but I guess there is no big difference.

Tim: No?

Miku: But there is a big difference, so at that time, my friends told me I was crazy to start my own startup when I was a student, but now, most of the students want to start a business, especially in my department because my department, for example, CEO of Packcha Technology and CEO of Gunosy graduate from that department and so they ate, for example, ramming together and they became super rich, right? So, they feel yeah, I can do that as well if he made it.

Tim: Role models.

Miku: Yes, exactly.

Tim: Role models are so important.

Miku: Exactly.

Tim: You know, to be able to look at someone who is not that much older than you and like, yeah, I could do that, I could do that.

Miku: Yes, so they changed a lot, and also more than 10 years ago when I was a student, it was super difficult to raise funds, so at the time, I did $1 million, series A round and it was a very big deal at the time, but now, $1 million is…

Tim: Is small, yeah. It’s crazy, isn’t it? But yeah, especially at the University of Tokyo, they’ve got a really strong fund and very supportive venture environment there.

Miku: So, it is great for the Japanese startup ecosystem.

Tim: I think it’s fantastic. It’s exciting to watch how fast it’s changing.

Miku: Yeah, it is surprising.

Tim: Yeah. Well, listen, Miku, before we wrap up, I want to ask you again my “Magic Wand” question which is if I gave you a magic wand and I told you that you could change one thing about Japan, anything at all — the education system, the way people think about women founders, the way people think about risk — anything at all to make it better for startups and innovation in Japan, what would you change?

Miku: So, I have read a report before and there is a big correlation between the percentage of startup founders and the percentage of GDP growth. There is a very strong correlation and if the percentage of startup founders increases 1%, it impacts 0.5% of GDP growth.

Tim: Wow.

Miku: So now, Japan’s GDP growth is 2%.

Tim: Pretty slow, yeah.

Miku: Very little, right? But if we can increase 2.5%, it is a big growth.

Tim: Just by getting more people to start up, to try.

Miku: Yes, so AI will change that.

Tim: I think we are already seeing that. I mean, I think every year, more and more startups are getting founded.

Miku: Yeah, definitely, it is increasing.

Tim: Yeah, and I hope we just see that trend continue.

Miku: Yes, I hope I can be a role model as well.

Tim: I think you already have been, I really do.

Miku: Thank you.

Tim: Listen, Miku, thank you so much for coming in.

Miku: Thank you so much for having me here. It was great to talk to you again.

Tim: Likewise.


And, we are back.

Now, there are two things that came up in our discussion that I really want to dig deep into. First, I think it’s amazing how much support Miku got from my clients when she was pregnant.

Now, for those not familiar with Japanese business protocol, because of her age, the size of the companies, the fact that she was the vendor, I mean, every bit of Japanese business etiquette dictated that she must visit her clients at their offices. The fact that many of her customers were willing to overlook the traditional rules to make things a little bit easier for an expecting mother is awesome.

It’s encouraging but I want to temper that encouragement a little bit. Women in Japan still have a rough going in traditional companies and these kinds of accommodations are not generally made. I think the reason for the discrepancy might be that enterprise Japan views startups as somehow fundamentally different from traditional businesses and they are. I mean, the CEOs wear T-shirts, women lead programming teams, there are no set office hours, but no one is exactly sure what the rules are supposed to be, so people or at least nice people who tend to be more accommodating.

Miku’s potential customers probably would have treated her very differently if she had been in sales at, say, Mitsubishi, but still, this is a great sign, and with all the abstract talk about making corporate Japan more friendly towards women, it’s nice to see a concrete example.

Second, Miku’s comments that Japan is actually ahead of the US and China in business AI goes against everything I’ve been hearing and reading on the topic. So, I did a bit of research and I think Miku has a real point here.

The boundaries of AI research are being pushed in China and the US to be sure, but most AI startups are using the same AI models and the same basic techniques, many of which have been around for decades, so in terms of who will have the most advanced applications, well, that’s going to be the startups with the best insights into their customers’ needs and those with the customers who have the biggest needs, and as Miku pointed out, Japan has a huge need for paper document reading.

Solving the complex and difficult problems of Japan’s corporate processes means that Cinnamon is more than ready to handle the comparatively streamlined corporate processes in the US and Europe.

And let’s just take a moment to appreciate the irony here.

The fact that corporate Japan has been so slow to modernize and its insistence on sticking to inefficient processes, that is the very thing providing a huge competitive advantage to Japan’s business AI startups.

If you want to talk more about AI or women founders, Miku and I would love to hear from you, so come by and let’s talk about it. If you leave a comment, I guarantee you that Miku and I or maybe both will respond.

And hey, if you get the chance, check us out on LinkedIn or Facebook, but even better, if you like the show, tell people about it. Disrupting Japan was recently ranked as the third biggest business podcast in Japan and that’s really thanks to you. Disrupting Japan is grown not by social media marketing or advertising but simply because listeners like you enjoy it and they tell their friends about it.

But most of all, thanks for listening and thank you for letting people interested in Japanese startups and innovation know about the show.

I’m Tim Romero and thanks for listening to Disrupting Japan.