JA

JP STARTUPSJP STARTUPS

The web magazine that introduces and supports Japanese startups

We are yet to witness breakthroughs with AI. Making society simpler through the social implementation of cutting-edge technology|Koichiro Tamura of ACES

Share:

This website has been translated into English using automatic translation. Please note that the translation may not be entirely accurate.

Since the third AI boom in the 2000s, research and development of AI has flourished, and the number of companies engaged in AI business has increased. However, did you know AI itself is said to be becoming a commodity? In November 2022, ChatGPT, a chatbot that realizes high-quality dialogue despite no code, was released to the public and made a significant impact. The enterprise value of OpenAI, which develops ChatGPT, is said to be 29 billion USD (3.7 trillion yen). In this era, AI can be handled with no code and low code by non-AI engineers as well. How can Japanese AI startups and engineers demonstrate their value amid calls for the commoditization of AI?

This time, we introduce ACES, Inc., which continues to grow by revolutionizing the AI development process with its platform. What is the future that ACES aims for, and why did it decide to develop it in the current format? We interviewed CEO Koichiro Tamura about various aspects of ACES.

"Algorithms make society simpler" | ACES' challenge that began with manga translation

What inspired you to start your own business?

The impetus for starting my own company came when I researched AI at university and had a vague idea of what I could do for society with AI..

In a paper published in 2017, a model called Transformer was proposed, which led to a breakthrough in the field of natural language. Around that time, while taking a shower at the hotel I was staying at for an academic conference, I talked with a friend about how we could do something interesting using Transformer's technology, and we came up with the idea that it would be interesting to use AI to translate Japanese manga and deliver them to the world.

After that, I had the opportunity to talk with a publisher I knew, and I realized that I needed a corporation to carry this project forward, so I launched ACES. In the end, the original manga translation project fell through, but I thought, "There are very few people who can implement deep learning in society. I want to bet on the potential of deep learning with this group because we might never get another chance to gather such a group of people". From there, I turned to the DX business and laid the foundation for our current business.

How did you develop your strategy after the pivot?

First, I thought about an AI startup's winning pattern. Data is important not only for startups but also for AI development, so in general, it would be necessary to partner with large companies with a lot of data. However, if you merely outsource from them, it will no longer be original. When looking into the business model, I thought that "a business model that looks like an outsourcing project but isn't outsourced" would be sustainable. Therefore, to provide them as modules, we started developing semi-custom AI algorithms and licensing them to large companies.

Why did you choose to provide them as modules?

Credit: ACES, Inc.

At ACES, we have developed the ACES Platform, where modules are accumulated and structured so that the more system development projects you carry out, the more algorithms you accumulate. We have achieved scalability and faster development by applying, enhancing and adapting the algorithms developed in previous projects to the next project.

This development process is also relatively cost-effective. Even for new development projects, we have modules developed in-house and stacked with the high level of AI we have accumulated so far and can assemble them like blocks as needed. Since there is no need to develop algorithms from scratch for each project, high-profit margins can be achieved.

You've also focused on the sales AI tool ACES Meet.

yes. ACES Meet is a sales support AI tool that allows AI to record and create minutes of online business discussions and share and analyze the content and temperature of these discussions.

You can check the content of business discussions and meetings by taking minutes; however, it does not help in accessing actual communication, such as facial expressions, reactions, gaze, intonation, and speaking speed. ACES Meet saves business discussions as a video and analyzes the recorded data to visualize speech habits and behaviors that even the person may not have realized. There are various uses; for example, analysis of the facial expressions, gaze, speaking style, and content of top salespeople can be utilized with excellent sales know-how and applied to training new employees. The DX of meetings can also increase the value and find areas for improvement, making meetings more efficient. Especially in modern times of remote work, where the number of meetings is increasing, I want to promote the productivity of meetings with ACES Meet.

Please tell us how you developed ACES Meet.

Many businesses and operations are built on the knowledge and proficient skills of people, but there are many restrictions due to personalization. Deep learning enables the structuring and digital reproduction of knowledge and skills that are close to people but were previously challenging to organize.

The field of communication is another area that is built on the knowledge and skills of people. In addition, meetings and business negotiations were often held offline, but the COVID-19 pandemic has caused them to be moved online. This shift has created opportunities to digitize sales and meetings, which had been challenging to digitize until now. This area is becoming popular in the United States and Europe, and we developed ACES Meet to provide it as a product for this purpose.

Please tell us about your vision of "Algorithms make society simpler".

The vision is a question for us and why the company exists. One of the features of deep learning is that it can process complex things in a complex manner. Until now, only humans have been able to process things in a complex manner, making an individualized response necessary; however, I believe that deep learning algorithms will replace this in the future. In other words, algorithms can create a simpler society for humankind. This is the background for the vision of ACES.

And by eliminating individualized work with algorithms, it is possible to create margins. By using that extra space for self-growth and a little leisure time, you can expand your options in life. This vision was born from our desire to realize such a society.

You provide DX partner services, but many companies face challenges, such as a lack of human resources, dependence on vendors, and only ending up with PoC (proof of concept). How does your service aim to solve these problems?

We aim to solve such problems by providing three fundamental values.

The first is designing the business value of AI in detail. The purpose of DX (digital transformation) is not digitization but transformation. Rather than relying solely on AI technology, we design the business value of AI based on a clear understanding of the current situation and goals, listening carefully to what kind of problems are occurring in real businesses and industries and focusing on how we can use AI to solve them. 

The second is the ability to achieve both high accuracy and reduced development risk.As mentioned earlier, our core strength is a technological foundation that allows us to efficiently promote digitization by utilizing the technological assets we have accumulated through research and business.

And the third and final point is having the leadership to drive the business forward and the ability to make progress. These points are important as DX is not just a bottom-up improvement but needs the energy to bring about change in businesses and industries. For this reason, all our personnel can get involved and move forward with the mentality not to be daunted by obstacles and difficulties. Rather than ending with PoC and demonstration tests, we work with the customer to achieve their goals after conducting "AI value design" that will lead to a solid business.

Is the value of AI declining? What is required of the modern AI researcher/engineer?

With many algorithms available as open source and the number of services that can handle AI with low code or no code increasing, what are your thoughts on the commoditization of AI?

If you think about it in terms of people, abilities such as voice recognition, for example, are cognitive abilities that most people have. However, areas that require specialized knowledge and skills, areas with so-called experts, are difficult. For example, if you asked people to solve a question from the University of Tokyo entrance exam, I think only about 10% of people would be able to solve it. Such areas are equally challenging for AI. As far as what AI has commoditized is concerned, it has not yet replaced most jobs.

I believe that basic cognition, such as voice recognition that I mentioned earlier, or the things that animals do regularly, are becoming commoditized, but that does not mean that AI has become common in business. In particular, very few companies can adequately envision future business and goals that will generate value in the next 5 to 10 years.I believe that the extent to which AI can contribute value to the business will remain as large as ever.

What skills do modern AI researchers/engineers need?


In addition to the ability to understand and implement algorithms, we believe the speed of hypothesis testing is important. In particular, speed is of the essence for engineers as they repeat the cycle of making hypotheses, asking questions, and verifying each point.

I am talking as an R&D professional at ACES rather than as an academic researcher. However, I still think it is important to be aware of research that creates business value and to pay attention to the field.

You are conducting extensive development in AI. What kind of people work at ACES?

It's mostly people who like technology. At ACES, everyone is expected to be an ace in a particular field, and everyone is expected to demonstrate their expertise in their respective domain.

Besides, I think there are many people with a driving force. Although they are engineer-like, they are also rational, allowing them to push through with power when necessary, giving their best to achieve their goals.

How did you spend your time as a student?

When I was a junior high and high school student, I was skeptical and frustrated about various things, such as when I would use the study materials and why it was necessary to copy whatever was on the blackboard into my notebook. When I became a university student, I was freed from those rules and restraints and started taking on new challenges that I had never been able to before, such as starting my own business and becoming a trader with the money I had saved. In my second year as an undergraduate, I lived like a day trader, waking up before the market started, analyzing companies and news, writing programs to analyze market data, and analyzing other investors' Twitter feeds in real-time to make investments.

One day, while living such a lifestyle, I lost about 3 million yen when I took my eyes off the market for a few hours. The shock of this loss sparked my interest in financial engineering, and I joined the Matsuo Laboratory at the University of Tokyo. In addition to the strategic thinking I had from the beginning, I believe my interest in corporate analysis and corporate management at that time, combined with AI technology, led me to what I am doing now.

Do you have a message for startups in the pre-seed to seed stage?

I think it's important to increase your turns at bat, even if it's a mistake. For example, the idea that you can change the world with your products, or the awareness that you can do so, is half-overconfidence, half-misunderstanding. It's the same with research. Most of the time, someone has already done something and failed. Sometimes, you go to bat based on the mistaken belief that your research will change the world, but if you do it about 100 times, even from a mistaken belief, you may happen to score or hit a home run. I want to convey the importance of standing properly to bat, even if it is a misunderstanding, in both R&D and business.

Finally, please tell us about your future prospects and give our readers a few words of advice.

I have various thoughts while running the company, but I have a strong sense of crisis that "I have to do something about Japan first." Although there is rapid progress in the declining birthrate and aging population, many futilities and inefficiencies still exist in industrial sites. In addition, many businesses and operations are built on the knowledge and proficient skills of people, so there are many restrictions due to personalization. By incorporating AI algorithms into our business, we at ACES will create a revolution allowing humans and AI to collaborate, evolve and grow together.

I would be delighted to work with people who sincerely wish to work toward the realization of this transformation, that is, "AI transformation", or to work together as partners.

Ranking