EdgeOf is an important startup ecosystem player in Tokyo with a huge international network, global talent and an amazing office location in the heart of trendy Shibuya.

For the premier issue of their magazine "The Crossing", Kageyama Hiroyuki, the Editor in Chief, interviewed our CEO Clemens Wasner on a wide range of topics - from our company, its mission and goals up to the Austrian AI ecosystem and how we compare internationally. Below you can find an abridged version of the interview:

Please describe EnliteAI and its services?

The best way to describe EnliteAI is by imagining a combination of a specialized management consulting firm and an AI Lab, where we are pushing the boundaries of AI by implementing and optimizing the latest AI research & scaling AI to industrial needs. This combination enables us to support our clients along their entire AI transformation – from problem identification and strategy analysis to implementing AI-based solutions, e.g. fraud detection, autonomous driving, video generation and so on.

What was the motivation behind starting an AI company in Europe?

The idea for starting such a company came during my time in Asia from 2006 to 2016, when the current AI hype began and autonomous driving started to make headlines. It struck me that Japan and China took AI very seriously from the beginning (around 2014), whereas European media and companies thought of AI as just another toy.
For example, Toyota invested $1bn in an AI lab in Silicon Valley already in 2015. At that time, German carmakers were still trying to figure out if AI is worth being invested into. Therefore when I returned to Europe, it was clear for me that it is necessary to guide companies on their journey towards AI.

What is your business focus and how do you differentiate from other companies in that space?

We are neither offering cheap implementation services like front-end or backend development, nor project management services. It's all about creating business value with AI.
In order to attract top talent, which is scarce within the field of AI in Austria, we do plenty of our own R&D. We call this our internal "AI Lab".

I'm sure you are familiar with NVIDIAs project to let AI create faces through AI. We applied this methodology to create our company logo.

The method we used is called GAN (generative adversarial network). The basic idea is that you have 2 competing neural networks:
#1 the generator, who wants to create life-like images.
#2 "the discriminator", who checks if the "fakes" from #1 are lifelike or not.

For our logo, we downloaded >100,000 Logos from Crunchbase and afterwards trained a GAN for 2 weeks on it.
The results are probably not amazingly creative, but it got the job done.

There are very few companies with experience in applying GANs to real world problems.

When German carmakers were looking for companies with that expertise, they found us. You could say an internal "toy project" opened us the door to a billion dollar company.

I think this also exemplifies the state of AI in the business world: It's still very early, there is a lot of experimentation and not so many blueprints and best practice examples. This allows a small company such as ours to out-navigate big companies.

What do you think are the main reasons some companies fail to fully exploit the potential of AI technology?

Why companies fail? Either they got too high or too low expectations.

Let me explain that a bit more in detail:
Too high: previous technology waves (e.g. big data or smartphones) created the expectation that a company can start with an all-encompassing strategy and afterwards goes into implementation. This doesn't work with AI, where we are still in the early days. So instead of starting with a big strategy, we recommend our customers to first get some experience with what AI can do for their business, e.g. through prototypes or proof of concepts.
Too low: with "too low expectations" I'm referring to the expectation of companies towards themselves. Companies look at Google and Microsoft and often think: "These companies have so much data, we will never be able to compete with these giants since we are too small." This creates a self-defeating mindset, which I unfortunately often encounter. Once you think of your own data as "too low", it paralyses your ideation process.

Do you believe that many companies can still create new business value from AI, even though large amounts of data are owned by tech giants in the U.S. and China?

To answer the first part of your question: Tech giants have lots of data indeed, but this is limited to a relatively small area, mostly internet and eCommerce usage:

• who clicked on which advertisement
• people who bought product A also went to buy product B
• customer sentiment in product reviews and chats, etc.

When you look at the European and Japanese economy, there is still a very strong もの作り element, which just doesn't play a role for Google - in other words, our economies sit on large amounts of data related to the real world (as opposed to internet data) which we can use to create entirely new things.‍

Toyota and Volkswagen have more data on production and supply chain than anybody else in the world, fuse this with AI and you have a similar advantage as Amazon has with eCommerce.

‍

Full interview in The Crossing, Issue #1: Link