[THE INVESTOR] Placing big bets on artificial intelligence, tech giants like Google, Microsoft, IBM and Samsung have been making huge investments, rolling out their own versions of AI-based assistants.
So far, they have mostly been in the form of digital assistants that understand natural language queries. The products in the market, however, are not sophisticated enough to cater to the real needs of people.
Ted Cho, CEO of Skelter Labs
To take the next step, and prepare adequately, it’s all about securing core AI technology that can fully understand context and address individual needs, says Ted Cho, founder of Seoul-based AI tech firm Skelter Labs.
Cho previously served eight years as the R&D chief of Google Korea. He also founded Saerom Technology, a stock market darling in Korea once, and developed Dialpad, the world’s first free internet call service, which was later acquired by Yahoo.
He warned that the Korean tech sector, which lacks skilled experts in AI and looks down on engineering, may end up struggling to catch up with front-runners.
“(When I was working at Google) I literally looked at every single Korean startup for potential acquisition, but few were technology-driven,” Cho said in a recent interview with The Investor.
However, it’s also true that many technology trends in Korea have been underappreciated since the dotcom bubble burst in 2000, and commercialization and profits are often given more priority than a startup’s growth potential and technology, he added.
“Whether a startup has the essential technology plays an important role for global investors’ investment decisions, but that is not the case in Korea,” he said.
Skelter Labs is Cho’s fifth startup consisting of seasoned developers and engineers from Google and KAIST graduates and aims to roll out game-changing AI technology that provides services tailored for individuals, way ahead of existing platforms.
The following are the excerpts from the interview.
The Investor: What does AI mean for Skelter Labs?
Cho: AI is a life-changing opportunity for Skelter Labs. I think our current projects will be worth millions of dollars in the future. However, the numbers are right now meaningless. Our main priority is developing core AI technology. Money is secondary.
TI: What’s your take on the recent AI boom?
Cho: AI should not end up as a short-term fad. It can change our lives in the long term. What worries me about the recent AI enthusiasm is the term, AI, is used in a reckless manner. Any company which makes use of the AI technology calls itself an AI firm these days. That is actually incorrect. The firm that actually works to develop AI technology should be labeled as an AI firm, not those simply using the technology.
TI: How would you define Skelter Labs’ vision in one statement?
Cho: Skelter Labs has been going eyes out to develop AI technologies that are innovative enough to change the world. Our ultimate goal is to develop a new generation intelligent personal assistant platform that goes beyond the limitations of current and traditional ones. Our tech focus is to create this platform with four main key AI advancements: personalized, anytime and anywhere, context aware and capable of natural emotional conversation. Aggressive investments and superb engineering talent are essential to achieve this dream.
TI: How will you target the global market?
Cho: Skelter Labs has been collecting data of human behaviors such as emotions, context and preferences and analyzing them to develop technologies for the IPA that has a full understanding of the users, provides useful information and sympathizes with them. In short, I would say that Skelter Labs is aiming to create technology that can be universally applied in all markets.
TI: Could you tell us about yourself and the team?
Cho: I have long been associated with AI since my graduate days at KAIST where I studied AI for my Master’s and Ph.D. I’m also an entrepreneur. I founded Serome Technology about 20 years ago and created Dialpad, the very first free-of-charge internet phone service, which later was acquired by Yahoo. Afterward, I banged pipes in Silicon Valley, competed against global tech companies and learned valuable lessons. Starting in 2007, I have led multiple global projects as the engineering site director of Google Korea for eight years. While I was working for Google, I witnessed Korea falling behind in AI, and wanted to challenge myself in that area. A few colleagues came onboard, and that’s how Skelter Labs began.
Skelter Labs consists of top brains with various backgrounds in AI such as Google and KAIST AI Lab. In addition, 70 percent of our engineers have a Master’s or Ph.D in computer science. Besides our engineering excellence, we have a great mix of product managers and UX designers as well.
Skelter Labs' Seoul headquarters
TI: How much funding has Skelter Labs raised so far?
Cho: As you may know, KakaoBrain and K-Cube Ventures have invested in Skelter Labs this year. The exact amount of each investment cannot be disclosed. However, we continue to explore new opportunities to get closer to our goal.
TI: Is Skelter Labs collaborating with KakaoBrain to develop AI apps?
Cho: Skelter Labs has been doing R&D projects on its own so far. The two companies haven’t yet discussed exactly when and how to cooperate with each other. However, we will continue to build a positive relationship and see how the partnership goes in the future.
TI: Identify Skelter Labs’ rival and your competitive strategy.
Cho: Some companies claim they have developed an innovative AI solution which turned out to be a mere voice recognition speaker or a simple chatbot. Skelter Labs, on the other hand, is looking at the bigger picture. What we are aiming to do is to develop technologies and mediums such as context recognition, recommendations and emotion modeling rather than simply utilizing voice recognition to allow users to enhance the quality of life. In that sense, it’s difficult to define who our main competitors are.
TI: Please tell us about apps you have already launched, or those under development.
Cho: We have been analyzing daily use-cases of IPA on different platforms such as smartphones, home appliances, cars and wearables and predicting which technological approach can be applied in the future to create a ubiquitous IPA that goes beyond the limitations of traditional voice speakers.
We’ve developed AI-powered applications such as those for flight search, and movie and restaurant recommendation, some of which are still in the pipeline.
Right now, we are working on four projects. First, a personal assistant that provides personalized recommendations at the right time and in the right place. Second, a scalable and conversational chatbot platform that can be applied by businesses in key industry verticals. Third, a deep-learning project for building AI technologies that match high-level cognition. And lastly, a new type of portable IPA device that integrates all of our technologies.
TI: Can you tell us about Google and why you left?
Cho: Google was certainly different from other companies. It puts high value on innovation. When I joined in 2007, I was surprised to see a firm of that size move so fast, almost like a startup. There was a sense of freedom, no set fixed rules, and most things were motivation-oriented. I once asked Eric Schmidt (chairman of Alphabet) about the budget for Google’s Korean office and he said “sky is the limit.”
Even though Google is heaven for engineers, its projects are usually so big and individuals can seldom have their ideas approved by decision makers. When I offered a business plan that I assumed could generate hundreds of billions of dollars in sales in the first year, a top exec at the Mountain View office rejected it, saying “It is a good idea, but not worth trying,” since the California-based firm makes the same amount of money a day. Then, I realized I had to leave Google to start an innovative startup on my own.
TI: Why do you think global investors are not interested in investing in Korean startups?
Cho: The first thing I tried to do at Google was take over Korean startups. I had always wanted to pound the pavement for the M&A market for startups. However, I was shocked that there are very few technology-driven startups. Even if a startup has great technology, the portion of engineers remains a mere 20 to 30 percent of the entire workforce and the rest are salespeople and marketers. These kinds of companies cannot be attractive from the investors’ perspective. In M&As, investors buy technology and products. Having too many non-engineers means investors should conduct large-scale layoffs after an acquisition which they certainly do not want to do.
By Kim Young-won (email@example.com)