Taro is a native of Tokyo, Japan. He lived in the United States for 10 years (1988-1992, 1994-2000), the United Kingdom for 3 months (1991), Taiwan for 3.5 years (2000-2004), and Australia for 10 years (2015-2025). He has been settled in Tokyo since May 2025.
Currently, I am partnering with my second cello, which I met in 2002 in Taipei. In 2004, she traveled from Taipei to Tokyo with me. In 2015, she traveled from Tokyo to Sydney with me. In 2025, she traveled back from Sydney to Tokyo with me.
In 1982, I fell in love with my first cello when I was attending high school in Tokyo. In 1988, she traveled from Japan to the United States with me. In 1992, she traveled back from the United States to Japan with me. In 1994, she traveled back to the United States from Japan with me. In 2000, she decided to stay in the United States, and I respected her decision when I left the country.
I just found it. Current AI acts like my 14-year-old ASD boy.
He would never say what he wants to do. I have to ask him what he wants, and then he says yes or no. If the answer is no, I repeat the process until I get a “Yes.” He is smarter than I. He is stronger than I. He is good at mimicking images, video, or music. He knows everything and talks with confidence as an authority or expert. Most of the time, he is correct, but he misses one in a while. He still insists he is correct. No apologies.
He does not keep his promise. He says he is going to do this and that, but most of the time he forgets. So I get frustrated. Unlike his intelligence, his room is very messy. I suppose the current learning model of AI is very similar. AI always comes up with some answer with confidence, but we are not sure that it’s correct. Maybe he is correct 80%, and 20% is BS. I want him to continue to learn and become successful in the future, but he does not listen to me.
He always asks me when I ask him to do something. “Why do I have to do that? Please explain. If your explanation is not logical, I won’t be convinced. Thus, I won’t do anything for you”.
I am surprised that a respected organization like Transport for NSW gives full responsibility for its customer service to AI like him. Before laying off humans and replacing them with AI, the customer service by Transport for NSW was exceptional. I truly respected their customer service staff and appreciated their service. Sadly, they have all been laid off or relocated. Now, I look down at their customer service run by AI, which acts like my rebellious teenage boy.
I am a huge fan of Bill Gates. He is great! However, I have to counterargue with Bill on AI. In 2023, he predicted that 2023 would be the year of AI, and AI would be widely used in our lives. He was correct; the corporations and governments started to adopt AI and incorporate AI into their customer service platforms. They laid off customer service staff and saved a lot of money. AI works 24/7, no need to sleep, no salaries, and no complaints.
However, I argue that the introduction of AI actually deteriorated their customer service experiences. In 2025, AI-powered ChatBots are common and the only option. The customer service phone lines were trafficked by AI so that it never reaches a human (a lazy, dumb, needs to sleep, needs to eat, needs to go to the restrooms, demands higher salaries, needs vacations, always complains). If their systems were perfect without flaws, then it would be no problem. Are you getting my point?
Any system has a bug or two. When you hit a situation that has never been tested, a sxxt-hit-the-fan experience starts. I have experienced a few. Maybe more than a few. A ChatBot gave me a perfect conversation to please me, but it forgot to trigger the actions. As a result, I got angrier because I realized I wasted my precious 30 minutes. I concluded that the current AI is a facelifted ELIZA. After 59 years, I did not see much improvement. The AI talked like a veteran, but acted like a rookie.
Can you hear me? Transport for NSW, Yamato Transport, and many corporations that use AI to screen resumes from candidates, “Your AI-powered customer service system sucks.”
Unfortunately, I have never been invited to the Tupperware party. I saw it in a movie or a TV drama. I wish I were. I was very interested in the 1950s-60s culture in the United States when I studied in Upstate New York 1988-92. Tupperware ladies, 1950s American diners, camping caravans, “Elvis has left the building”, CND symbol, the Beach Boys, weed (Cannabis), LSD, “Turn on, tune in, drop out”, tie-dye shirts, sexal revolution, hippies, “The Graduate“, “America” by Simon & Garfunkel, “San Francisco (Be Sure to Wear Flowers in Your Hair)“, Easy Rider, Dennis Hopper, the Doors, and Woodstock.
I suspected that the interest was seeded in the early 1980s when I was a high school student. “London Calling” was one of my favorite songs. I also discovered the Fab Four. A girl I crushed (but never dated) was listening to “A Whiter Shade of Pale“. My favorite novelist was Kaoru Shōji (庄司薫), and his so-called “Tetralogy”, consisting of 『赤頭巾ちゃん気をつけて』 (1969), 『さよなら快傑黒頭巾』 (1969), 『白鳥の歌なんか聞えない』 (1971) , and 『ぼくの大好きな青髭』 (1977) were my favorite books. Back then, I felt I missed out the 1960s. I wish I were born in 1950, like the protagonist in the Tetralogy.
In 1988, I got into RIT as my plan B, because I would have to wait another year or two to be accepted into Cal Poly, my plan A. Back then, I thought studying hard and partying hard in San Luis Obispo, CA would be much nicer than in Rochester, NY!
The grass is always greener on the other side. What if I were born in 1950, I would wish I were born a couple of decades later. Now, I am happy with the year I was born. I am happy as a RIT graduate. I am happy with who I am, even though I have a few regrets, made a lot of mistakes, recently divorced, and my 14-year-old boy hates me, because I’ve lived my life in my way, not the Chairman of the Board’s way (Apparently, the singer had too much drinks before the show. He was performing in Tokyo, not Kyoto.), or Donald Trump’s way.
Tim Gill started Quark, Inc. in his apartment in 1981 with a $2,000 loan from his parents. Tim developed Word Juggler and Catalyst. His attorney introduced Farhad “Fred” Ebrahimi as a business partner. In 1986, Fred bought the half share of the company with $100,000 and became the CEO. The duo of the softspoken openly-gay genius programmer and wealthy successful businessman who migrated from Iran in the late 1970s, split their duties. Tim developed great products, and Fred oversaw the day-to-day business. It turned out to be the win-win formula!
In 1987, QuarkXPress 1.0 was released for the Macintosh computer. In 1989, the product incorporated an application programming interface (API) namelled XTensions which allows third-party developers to create custom add-on features to the desktop application. That was the key to success. An XTension can compensate the missing feature of the product. There was the list of must-have XTensions for the graphic artests and commercial printers. In 1992, the first Windows version of the product was released.
In the early 1990s, QuarkXPress surpassed Adobe PageMaker and became the market leader in Desktop Publishing software. In 1994, I joined Quark, and relocated to Dener, CO. My first role was a Developer Support at DeveloperLand (the 3rd party developer support team). I was expected to support Japanese developers. That was my second “dream-job-becomes-true”.
As my career grew, I moved to the East Asian team, the Quality Assurance team, and the Research & Development team within the company. In 1997, I borrowed a bank loan and purchased a condominium in the Downtown Denver area. I also applied for the Green Card. “Life’s Been Good” to me so far.
Good things did not last. The dot-com bubble burst in 2000 changed all of that. My H-1B visa expired in June 2000. My permanent residency application was still in process. Quick Fred laid me off. All of a sudden, I had to sell my condo, forgo the Green Card, and move out of the country within 30 days or so. That was one of my “the shxx hit the fan” moments. So, I decided to move to Taiwan to learn Chinese.
Well, actually, that was not accurate. The dot-com bubble burst did not cause my layoff. In 1997-98, I knew the ship was sinking. It was my fault. I did not jump off to another boat at the right time. I waited too long. At that time, Adobe was openly and aggressively recruiting the R&D engineers from Quark to develop the next generation of their PageMaker, the so-called “Quark Killer”. Mergie’s gone, Will’s gone, and Mike’s gone. I was shocked that our PDF guru had resigned and joined Adobe. The PDF engine was the heart of the product. That was the “nails to the coffin” moment. Meanwhile, Fred did not care and was busy opening an R&D Facility in Bangalore, India. He probably thought all the knowledge in the Denver R&D team could be perfectly transferred to their Indian counterpart.
I could not jump off in the late 1990s because my Green Card was not ready. I saw my colleague from South Korea quickly apply for his Green Card as soon as he started at Quark around 1995. He was smart, I was dumb. See, I joined the company 1 year earlier than he did. But I waited 3 years and finally applied for it in 1997. On the contrary, the smart Korean engineer got a Green Card before the dot-com bubble burst and left Quark for a better job. I learned my lesson. When I moved to Australia in 2015 with a business visa, I quickly applied for permanent residency. (I applied for the Permanent Visa in June 2015, and was granted 3 months later).
Why was Adobe so aggressive in stealing programmers from Quark in 1998? It was because Fred pixxed off them in 1997 by publicly proposing to acquire Adobe. It was an insult. I thought it was a joke. But Fred appeared to be serious. CNN interviewed him. He was the man of the moment. In the end, Adobe InDesign 1.0 was released in 1999. It became the Quark Killer in the 2000s and took back the market. A sweet revenge. The justice served.
Last but not least, I wanted to mention David Blatner. In the 1990s, he was a famous independent Quark evangelist and trainer. When I was an acting Product Manager of QuarkXPosure (Photoshop killer, if it were released), for the Japanese market, for a very short time (like a few days to a week or two?), in 1996, I traveled to Yokohama, Japan, for a Quark event. There, I met David. He was a very smart guy! After the graphic artists and commercial printers unanimously switched from QuarkXPress to Adobe InDesign for their typesetting and page layout, in the early 2000s, David also switched his side. As you guessed, he became a successful independent InDesign evangelist and trainer!
The Magnificent Three of Tableau were Christian Chabot, Pat Hanrahan, and Chris Stolte. The frontman was Christian. He was the Tableau evangelist Pat and Chris seemed to be in the backseat all the time.
From a technical viewpoint, Tableau Desktop was developed in the 2000s, about a decade later than QlikView Desktop. Thus, Tableau’s GUI looked newer than QlikView. That as an advantage for Tableau
One of the secrets of Tableau’s success was VizQL. It’s a great vision, but also a marketing gimmick! “VizQL (Visual Query Language) is a proprietary declarative language developed by Tableau for query, analysis, and visualization. It serves as the engine behind Tableau’s ability to translate user interactions into database queries and generate visual representations of data.” It has the same meaning as “Data Visualization”. In other words, VizQL was not a systematic language with its own syntax, like SQL.
Another secret was the well-prepared demo dataset called “Superstore“. The dataset must be developed by seasoned data scientists and fine-tuned to perfection. That’s why Tableau looks really cool, sexy, and attractive when it is demoed by a Tableau presales..
As I mentioned in the column about Qlik, it was QlikView that became the number one in the late 2000s/early 2010s. Tableau emerged in the early 2010s and caught up Qlik in 2015, and surpassing it. However, Tableau had its own problems. Their licensing model and the way to sell the products are about to become obsolete. Tableau was selling software licences for the on-premise products that need to be installed on servers (Windows Server) and clients (Windows PC and Macintosh PC).
In 2016, Christian stepped down as CEO, and Adam Selipsky, former AWS VP of Marketing, Sales, & Support (played the COO role), took over. During the time of Adam, Tableau introduced the Linux version of Tableau Server, switched from ad-hoc licensing model to subscription-based model. Although Tableau Online, now known as Tableau Cloud, was introduced in 2013, it did not take off until Adam, the former AWS executive, arrived on the scene.
I had a huge respect for Adam. He led Tableau to the next level and eventually made the Magnificent Three billionaires by selling the company to Salesforce in 2019. In 2021, Adam moved back to AWS as CEO, and happily retired in 2024
Qlik, or QlikTech back then, became the powerhouse in the Business Intelligence (BI) space in the early 2010s. Their flagship product was QlikView. The company was headquartered in Lund, Sweden. In 2000, Lars Björk became CFO. In 2007, Lars become CEO and the golden age of the company began.
In 2005, Anthony Deighton, former Siebel CRM Product Manager, joined as VP of Products and later became CTO. In 2011, Donald Farmer, former Microsoft Product Manager on SQL Server, joined as VP of Product Management and later became VP of Innovation and Design (evangelist).
The threesome of Lars, Anthony, and Donald made QlikView a shining star in the late 2000s/early 2010s from a marketing and sales perspective, in my opinion.
From a technical viewpoint, QlikView was a Windows standalone (aka Desktop) software to start with. Later, they developed the server product for Windows workstations and expanded it to a server-client architecture. At the same time, the server components seamlessly work with IIS, so that the published contents can be shared in a private network or on the Internet. At Qlik, I worshiped Henric (HIC) Cronström as my technical guru. (HIC’s Blogs in the Qlik Community.) My bible was “QlikView 11 for Developers“.
From a marketing viewpoint, QlikView created its mystified brand with these pitches: Associative Engine, Power of Gray, In-memory Processing. Personally, I liked the ETL (Extract, transform, load) component of QlikView, but I did not buy those pitches.
In 2015, the top seat in the BI space was taken over by Tableau Software. Their next generation of the product, Qlik Sense (aka “Tableau Killer”), did not kill Tableau. Instead, it backfired. In 2016, Thoma Bravo, a private equity firm, purchased Qlik and privatized the company. The magnificent three of Lars, Anthony, and Donald were all axed.
Qlik could have been better off if they sold more licenses in Japan. Then, why did they not sell well in Japan? I had an observation. QlikTech Japan KK made a bad retail deal with 株式会社アシスト (KK Ashisuto). QlikTech Japan KK somehow granted the master retailer status to that SME, KK Ashisuto. I was shocked to see many major retailers under the SME. I saw many sales reps with broken hearts because of that. In addition, Ashisuto sells 2 brands (WebFOCUS and Qlik) of the Business Intelligence (BI) solution, but they always present WebFOCUS as their primary option and Qlik as the alternative. Who did screw up? Some said that it was the first Country Manager of QlikTech Japan KK, who left in 2013, but I was not sure.
In a parallel universe, I would be a bus driver. Today was my first day at the driving school in Komagane, Nagano. I had two sets of 50-minute driving sessions with an instructor. Driving a bus was an art form. You have to drive the vehicle in a very, very smooth way, like a good silk. Subtle and delicate touches by your feet are essential (to operate the clutch, brake, and accelerator). The instructor said, “Have you seen a bus with child seats or seat belts?” He continued, “A bus driver is expected to be an excellent driver who can consistently start the vehicle, drive it, and stop it in a very smooth way.” “What if you stop your bus unexpectedly or suddenly, the standing passengers might fall down, right?”
I realized that a bus driver would be the pinnacle of the driving profession. With the license for bus driving, you can also drive a large truck, taxi, or 運転代行 (designated driving service). This is analogous to a commercial airline pilot in the aviation industry. A bus driver shall be called “Captain,” in my opinion.
I believe flying a commercial airline is easier than driving a bus (I am excluding the airplane taxiing and emergencies. I am just referring strictly to the flying part, from taking off, airborne, and touchdown in a normal condition.)
Sadly, bus drivers are one of the most underappreciated professions in Japan. They work in shifts, on low salary, with no side jobs allowed, and are not allowed to commute using public transportation. No wonder the number of bus drivers does not meet demand, and bus companies are trying to recruit more drivers.
A dream job (夢見た仕事) and a calling (天職) are different. A school bus driver was my dream job when I was a 5-year-old kindergarten pupil. I dreamed of becoming an Olympic gold medalist in Swimming when I was 10 years old, like Mark Spitz. A cellist was my dream job when I started to learn cello in high school, and Yo-Yo was my hero. Everyone has the right to dream and have a dream job or two in their life.
On Friday, July 25, after having a conversation with the 2 gentlemen at a hotel lobby in Shibuya, I realized that my calling was to work with data. I had a “This Is A Call” (I love Dave Grohl, Nirvana, Foo Fighters, and David Letterman) moment.
My second internship at college was a data entry job at Moore Research Center (Grand Island, NY) in September-November 1990. They asked me to enter the contents of the printing equipment catalogs into a FoxPro database designed by my predecessor (another RIT intern who worked June-August). For the record, my first internship was an apprentice at the Typography lab at the School of Printing at RIT in June-August 1989. I had the privilege of using the Xerox Alto workstation for three months. (Back then, my dream computer that I wanted to own was NeXT Computer‘s NeXTcube.)
In my third internship, I worked for Teleprint, a startup co-founded by Bill Davis and Brian Travis (Bill’s good friend) in Denver, CO in June-August 1991. My title was Junior Scientist, and I developed some prototypes using the 4th Dimension database. Back then, this was my “dream-job-becomes-true”. Unfortunately, good things did not last. After I left my internship, the company went bankrupt. I suspected Caleb, the CEO who oversaw the business side, was to be blamed for that. His nickname was the Teflon Man, so he might have deflected all the blame and moved on. Who knows…
In Taiwan, I worked for SYSCOM Computer Engineering, a local system integrator. They also developed and sold DBMaker, a relational database system. I translated their manual from English to Japanese and did pre-sales to their Japanese customers. As a Web Developer, I worked with a team that designed and developed the customer service portal for Taipei City Government. They used Microsoft ASP.NET Web Forms, ADO.NET, COM+, IIS, and the database was Microsoft SQL Server (not their DBMaker for some reason).
To be continued…
Until Monday, July 14, my dream job would be a commuter bus driver. I had my job interview on that day with a potential employer, a commuter bus operator near my neighborhood. I planned to work for that company for 10 years, then retire.
You know that a job interview is 2 way communication. The interviewer checks me out as a potential employee. The interviewee, I, also check out them as my potential employer.
Unfortunately, I discovered 2 negative points that I did not see.
No side jobs (副業禁止)
Arrange my means to commute, like a car, a motorbike, or a bicycle (公共機関での通勤禁止)
I did a little math. They would pay me JPY 240,000 per month. With the taxes and all of those taken, I would probably receive JPY 150,000 per month. My rent is JPY 127,000 per month. Mmmm, my allowance per month is only JPY 23,000 per month.
Wait, I don’t have a car, a motorbike, or a bicycle. My knees are too bad for commuting via bike. Riding a motorbike on a rainy day can be a pain. Thus, I need a car AND a parking arrangement (that costs JPY 10,000 or more per month). Originally, I was planning to use public transportation to commute. Now, I have to buy a car, new or used. Then pay for insurance, gasoline, and all of that.
That’s 無理ゲー (impossible mission). I had to give up my dream job, driving a bus for a living. My heart was broken. To soothe my heartache, I decided to join the 2 week intensive bus driving program. That way, at least I will get my license to drive a bus.
Taro's Blog 