Dallas Startup Week: “scaling without stagnating” for corporate firms

It was a pleasure and honor to represent Thomson Reuters at the panel on “scaling without stagnating” as part of corporate innovation. The event was hosted by Capital One and sponsored by Dallas Innovates as part of #dallasstartupweek. The panelists Dalia Powers (CBRE VP/CIO & panelist moderator), Sterling Mah Ingui (Head of Go To Markets Fidelity Labs), Scott Emmons (The Current Global CTO), Sean Minter (AmplifAI CEO), Charlie Lass (MIT investor) and myself Tarek Hoteit (Thomson Reuters Labs) took turns discussing leadership, people, organization, process/change management, and technology to support innovation in the corporate world. For me it was also an opportunity to let startups in Dallas to know about Thomson Reuters, Thomson Reuters Labs (http://labs.tr.com), and our community engagements in Dallas. I even shared my personal journey on a major transformation of a product as part of corporate innovation hoping to encourage everyone not to give up and do the same and more. We also answered questions from the audience such as how startups can interact with corporate (my answer: persistence is key but if someone from corporate is ignoring you, find another contact. Don’t give up)

Why 22 hour game-playing and listening to 68 episode podcast matters?

Thimbleweed Park is an award-winning point-and-click game released in 2017 as a tribute to similar pc and Commodore 64 adventure games in the 80s. It is created by Ron Gilbert and Gary Winnick, with design and development assistance from David Fox, Jenn Sandercock, and numerous other individuals. You control five characters simultaneously in a story that in itself contains multiple sub-stories, interact with a multitude of characters in the game, and solve the puzzles that you encounter. Each of the animated characters has very distinct personalities that you cannot forget even after you long remove the game from your phone or computer.

After spending more than 20 hours playing the game over a period of time (22hrs and 15minutes in my last attempt per game record), I was so marveled with the game that I search the net on its origin and found a podcast by the game creators that ran weekly for 2 years from April 2015 to April 2017 plus one more episode in April 2018. Most of the episodes are 15 to 20 minutes where the game creators discuss the progress of the game development. One episode each month is an hour long and includes answering questions raised by Kickstarter-backer of the game as well as those posting comments on the game’s website blogs.

In the last couple of weeks, while driving to or from work, I went through all the podcast episodes in chronological order instead of listening to my usual music. My curiosity for listening to the podcast started with no apparent reason besides wanting to know more about the game, but it quickly turned into something bigger. It wasn’t about how to build a similar game or a nostalgia to bring back memories of the 80s. It was about developing something innovative with a tremendous focus on detail while maintaining a set of rules and constraints that the developers themselves intentionally decided to bound themselves onto to stay committed to their original idea, and that keeps the game with an 80’s look and feel but using modern-day technologies. The developers used to developed games together twenty to thirty years ago including Zack Mackraken, Maniac Mansion, Indian Jones, Loom, and more. They were in their twenties at the time. Now they are in their fifties and sixties with a lot of life experience combined with their lifelong passion for adventure games and science fiction – your typical geeks that I personal align with.

From a technology perspective, the developers leveraged a lot of the traditional modern-day technologies, such as Git for code repository, Adobe Photoshop for the art and animation in the game, software development kits (SDKs) for Steam, GOG, Xbox, Swift, Nintendo Swift, OSX, Android, Windows, Linux, and more. They were also blogging and podcasting their process which was something unthinkable of in the 80s, working remotely and using Skype for communication, and were interacting with their game supports on social media channels. In the eighties, such technologies did not exist, of course. The limitations of hardware and software on old computers like the Commodore 64 generated a lot of creativity but bounded with can be done in terms of graphics, music, and 8/16bit computing power. Nowadays, games run on 64-bit computers and mobile phones, and can easily leverage augmented reality, virtual reality, and artificial intelligence technologies to maximize gaming experience without sacrificing performance or cost. The developers for Thimbleweed Park decided to do none of the cool new stuff.

Thimbleweed Park creators made sure all the graphics are limited to the 80s era, but they perfected all the details of the animated characters and the scenes. They added voice, special effects, and music that still makes you feel it is the 80s. They also expanded the game to support international speaking game players. They added innovative ideas including Kickstarter backers could record a voicemail which can be played back in the game when a character in the game dials a number from the phonebook. Other game backers would send some text of books which the developers then add to the library scene in the game. The game also includes entertaining puzzles such as ones that make use of the microwave or the air blower in some bathroom or some berries in the forest. Some puzzles are challenging but have an “Aha!” moment when some character makes a phone call while the other character does something else. I am not giving details because I don’t want to ruin the game for those that would like to play the game. While the game itself was so innovative and entertaining, the conversations between the developers in the podcasts were geeky sometimes but and most of the times were like normal discussions that can be taking place between friends in your backyard.

The recording of the podcast episodes started at the same time the game was getting developed and continued throughout the development stages. Like in any typical development process, things start with early ideas and preparation work. The creators do not disclose the details of the game in the podcast because they do not want to ruin the story before the games get published. It made listening to the podcast episode as adventurous as playing the game especially when I started listening to the podcast after the game was released and while I was playing it. Unintentionally, you would relate the actual game with what is being discussed in the podcast. Sometimes you would hope to find clues about the game even though the game itself as an option for you to ask for clues. Moreover, you would start to wonder how the developers are organizing their work – one is the artist, another is the developer, and a third is the project lead and is also an artist and a developer.

The game creators talk in the podcast, and we listen. They call out specific feedback received from their Kickstarter backers and discuss it. Hence, you feel users’ voices are added into the podcast. Sometimes they invite other members of the team working on the project. The episodes frequently include humor and nostalgia moments from the past. Even their personal lives would get shared on the call. As the episodes progress, they share concerns about release timetables, bug fixes, issues with platform rollouts, and different problems which are not uncommon in any project. In general, the episodes felt human and not some business production, so why all this important? The authentic experience of playing the game and listening to their thoughts and interactions throughout the lifespan of the podcast that ran in tandem with the game development made me rethink about how innovative thinking, natural human behavior, and organized execution can generate amazing results.

When the efforts by the developers were rewarding to its creators and its players while at the same time it was fun and challenging then why not take more of such lessons and apply it in corporate innovation. Key takes that I learned from Thimbleweed Park the game and the podcast:

  • Nothing is impossible but don’t be overly ambitious. Even though the developers and the artists were masters in their field, they did not go overboard with the project. They stayed faithful to what they wanted to do and did it exceptionally well.
  • They dedicated regular feedback channels for their customers. They allowed their users to suggest ideas but made the call whenever a draw occurred on which idea to pick. They also gave an option to their Kickstarter contributors at a specific level to be part of the game by letting them record their voicemail which you can listen to in the game if you dial the person’s name and extension located in the game’s telephone book. That is cool and original.
  • The podcasts were very entertaining and informative. They didn’t shy away from discussing their past or sharing their concerns about the game development progress. A great lesson learned about honesty and humbleness.
  • The game itself had lots of challenging puzzles but also include an option to get a hint. Just like any product, you always need some guidance when you get stuck.
  • The story of the game is long but not long enough to give up. There was still something in the game that pulled you to continue playing. Even when you leave the game for days and then return, you can quickly get back to the same rhythm before you left the game. The user experience is excellent. Design a product with an intuitive user interface that makes it easy for you to return to where you last stopped and can help you recall what you should be doing next.

I can think of other ideas that one can learn from playing such a game. But at the end of the day, I can summarize it all as: design something as intuitive as playing some game, have fun doing it and let the users enjoy what they are doing with the product, be honest, do not over commit but do not underachieve either. One last thing, play the game!

Game information is available at https://thimbleweedpark.com

ThimbleweedPark-CharactersAtVista.png

Note: images are copyright of Thimbleweed Park.

My Story with Pink Floyd

Sometime between ‘91 and ‘92 I accidentally saw Pink Floyd Another Brick in the Wall part 2 video on MTV – the scene where school kids turn into .. Loved the song, so I went and bought a cassette with only The Wall Part 1 album. Never thought to buy Part 2 at that time. I would then just forward the tape to Another Brick in the Wall part 2 right when the helicopter sound comes up. I stayed doing that for months until one day I thought of playing the whole tape from part 1. I loved it! Right then I went down to the local music store and bought part 2 of The Wall. For some time my world of Pink Floyd was only around The Wall. Never knew anything else existed until I encountered Wish You Were Here on the radio. But I did not know that Wish You Were Here was for Pink Floyd. I went to music store and tried to imitate the song in case the owner could recall which band was it. When the owner said it’s Pink Floyd and handed me the Wish You Were Here album, the whole world of music changed for me ever since. The Final Cut came next for me, and later Roger Waters Amused to Death. But my love for Pink Floyd took a more obsession when I started listening to Unmagumma, Saucerful of Secrets, Animals, Pipers at the Gate of Dawn, Meddle, and, my best one of all, Atom Heart Mother. Darkside of the Moon was nice, don’t get me wrong, but it was Atom Heart Mother that I had loved the most. I was told that the last album for Pink Floyd was the 1984 Final Cut and I assumed the band was over. I was so wrong.

Two years later, sometime in 1994, the song Keep Talking comes up on the radio. I thought to myself the sound is like Pink Floyd so what if Pink Floyd is back? Surely so! Pink Floyd released the Division Bell. I rushed to the music store, this time the store in Hamra Street. The owner said the shipment of album CDs is on its way shortly, so I parked my car next to the store and waited in the car for a couple of hours. Waiting for it was so worth it. Sometime after the Division Bell came the Pulse live album. Not my favorite compared to Atom Heart Mother and The Division Bell, but it was Pink Floyd!!

The music of Pink Floyd is so beautiful. The guitars, drums, the piano, the voices, the synthesizers, the words, the album covers, the movie, and more. No offense to David Gilmour or Roger Waters solo careers, the preWall era with the team as a whole is the best. I thank Nick Mason for putting the Saucerful of Secrets concert. Even though it is not the Pink Floyd band, but his passion and drum talent is Pink Floyd. His concert Saucerful of Secrets is a Pink Floyd concert to me, and Pink Floyd is Pink Floyd forever! Now off to see Nick Mason Saucerful of Secrets concert here in Dallas !

Minecraft or Fortnite for your career?

A common past/present/future dilemma in the professional space is what I learned from yesterday, which job is available today, and what to learn for tomorrow. It gets more complicated as career and family building kicks in, and even more stressful as a person’s age. If you can imagine Past, Present, and Future as building blocks stacked on top of each other, Future bottom, Present in the middle, and Past on top, and the height of each set equals the years of experience, things get interesting.

If you have been in the industry for a long time, the long years of experience make the Past blocks taller and heavier and put more weight on Present and Future underneath. The set of Future blocks tend to be less in quantity, and the set of blocks that stand or break the stack would be the Present set of blocks. Hence, things get fragile….

The converse when you are in young in the field, your set of Past building blocks are much less in size than the future ones; your Present blocks have a tougher situation. The first set of Past blocks would not have enough pressure over the Present blocks, and the Future blocks are numerous, and, subsequently, harder to pressure.

Either way, the Present blocks are in a sensitive situation – pressure or not pressure from Past blocks, and break or make the Future blocks. Furthermore, the middle set of blocks constantly form some kind of balance between Past and Future unless some external force or King Kong tumbles the whole stack apart.

So what do you do if you want to an instructive Past, a less fragile Present, and a healthy stack towards the Future?

I don’t think I have the right answer, but I have a personal one instead. We can’t change the number of blocks that we stack as our experience throughout our lives from past, present, and future, but we can influence its quality. Think of playing constructive-building Minecraft rather than destructive-style Fortnite. Which type of blocks do you want to build – gold, steel, wood, shiny, cracked? How do you want them stacked? Horizontally, vertically, or both? Do you want to build a castle or a fort or a pyramid? Do you want something that is a work of art or science or a hybrid? Do you want to spend a lot of time making every block perfect or have fun stacking things? The ideas can go on and on… Last, play Minecraft as your career building strategy. Stop wasting time destroying things as in Fornite.

Lightning Talk – Course Sentiment Tweets with AutoML

I presented a 15-minute lightning talk on leveraging AutoML for sentiment analysis.

Quick and easy AutoML for Sentiment Analysis and Classification tasks

“Machine learning algorithms have evolved significantly in the last few years. AutoML is one of the latest advancements in the field that allows anyone to build and deploy AI products without requiring extensive knowledge in the field. The lightning talk will show case how one can build a production-quality sentiment analysis model using Google AutoML and Google Cloud with the least coding possible.”

I first showed case how to upload datasets directly into Google AutoML NLP portal and, from there, train a model and perform predictions. After that, I showed how I integrated the sentiment analysis model into analyzing Twitter stream using Django, Docker, Twitter API/Tweepy, Jupyter Notebooks, and PostGresQL, I published the code on GitHub under hoteit/courses-sentiment-tweets

Give or Take

There are two types of people. Those who give more than they take, and those who take more than they give. The ratio between give and take can vary from zero to infinity such as zero giving and all taking or no giving but all taking. I think that knowing where you stand in such a formula can make a huge difference in one’s own life accomplishments and everyday human interactions. Life nurture such as experience can dynamically and actively influence the change in ratio between giving and taking, while life nature, basically age and genetics, can yield an uncontrollable and more slower impact or hardly a change in behavior. In couple of weeks people will once again plan their new year resolutions. Do consider the giving versus taking ratio formula for the years to come. Hopefully it should be seriously more on the giving rather than the taking side, not 50/50!

AutoML – expert AI for the inexperts and businesslike

Nearly everyone around by now has either heard or used artificial intelligence (AI) and machine learning (ML) in some form or fashion. Some students are already publishing papers in the field while other students are applying various AI techniques in their research, internships, or just for fun. Professionals in the industry have either incorporated some form of AI/ML into their product or services or are currently considering it. Either way, AI and ML have a lot to offer but not without a good amount data, significant processing power, right skillsets, and a lot of patience with design and execution of such projects. For that problem, AutoML is a promising new technique in the field that allows researchers and professionals to make use of pre-trained models and cloud-based services to roll out AI solutions much more rapidly than building machine learning models from scratch. AutoML provides the methods and processes to apply, integrate, deploy, and scale machine learning intelligence without requiring expert knowledge. Major AI platforms, starting with Google and followed by Microsoft, H20.ai, and others are priming AutoML as the next evolutionary frontier in artificial intelligence so that humans can spend zero time recreating machine learning models from scratch, and, instead, focus on applying the models while letting machines take care of building them.

AI/ML supporting hometowns of international students: what, how and why?

Nearly everyone around by now has either heard or used artificial intelligence (AI) and machine learning (ML) in some form or fashion. Some students are already publishing papers in the field while other students are applying various AI techniques in their research, internships, or just for fun. Professionals in the industry have either incorporated some form of AI/ML into their product or services or are currently considering it. Either way, AI and ML have a lot to offer but not without a good amount data, significant processing power, right skillsets, and a lot of patience with design and execution of such projects. For that problem, AutoML is a promising new technique in the field that allows researchers and professionals to make use of pre-trained models and cloud-based services to roll out AI solutions much more rapidly than building machine learning models from scratch. AutoML provides the methods and processes to apply, integrate, deploy, and scale machine learning intelligence without requiring expert knowledge. Major AI platforms, starting with Google and followed by Microsoft, H20.ai, and others are priming AutoML as the next evolutionary frontier in artificial intelligence so that humans can spend zero time recreating machine learning models from scratch, and, instead, focus on applying the models while letting machines take care of building them.

References: -PwC 2017 report “PwC’s Global Artificial Intelligence Study: Exploiting the AI Revolution” https://www.pwc.com/gx/en/issues/data-and-analytics/publications/artificial-intelligence-study.html

-National Foundation for American Policy (October 2017) “The Importance of International Students to American Science and Engineering” http://nfap.com/wp-content/uploads/2017/10/The-Importance-of-International-Students.NFAP-Policy-Brief.October-20171.pdf

Chess and Coding

Even software professionals can be rated like chess players. Whether it is successfully delivering a complex project or getting rewarded based on an innovative solution, the pressures of keeping up with one’s own rating even if was something virtual can lead to self-imposed stress and fear. When a chess player is facing tough competition in a rated game, the fear of loosing that leads to a decline in rating can be overwhelming. But if the game is treated as unrated then ones’ rate would not be impacted by the win or loss. Stakes will be lower and the pressure would be less. This can apply to software developers as well. Occasionally pick up an unrated project, such as something of your own choosing or a new type of code you always want to try but were afraid to get measured by. Do it under your own pace. Once you feel more comfortable with it take a similar project as a rated challenge at your job or in your community. You will then feel more confident and more comfortable with what you do. That’s what I do when I either code or play chess. #technology