[Interview] Oh Jonghwan, CEO of Lab2Eye, "AI Analysis Clearly Has Advantages in the Cryptocurrency Market"

[Asia Economy Reporter Lee Jung-yoon] The Luna crash incident has brought volatility and investor protection to the forefront of discussions in the cryptocurrency market. Ordinary investors often buy coins based on someone else's words. In a cryptocurrency market where it is difficult to foresee even a short distance ahead and there is no reliable information to refer to, they press the buy button influenced by what they hear.

Cosign, an open platform launched by Oh Jong-hwan, CEO of big data analytics specialist Lab2i, provides information that can serve as a guide for investment decisions. Deep learning-based AIs called mentors present investors with real-time forecasts of specific cryptocurrencies’ expected prices, returns, and accuracy rates. CEO Oh explained, "AI has the advantage of quickly analyzing vast amounts of data that humans cannot handle," adding, "Cosign’s purpose is to suggest investment methodologies and help investors make the final decisions."

The AIs analyze various data sources such as global macroeconomic indicators, news and tweets, and charts to derive information. The AI examining global macroeconomic indicators uses materials like gold, oil prices, exchange rates, and stock indices. There are also AIs that analyze numerous news articles and tweets or identify chart patterns. Based on their analysis, these AIs provide signals consisting of entry price, target price, and target return for cryptocurrencies over specific periods.

CEO Oh has been a data analyst at Seoul National University’s Graduate School of Convergence Science and Technology since 2014. He has also conducted research on robot journalism. Over time, he accumulated know-how through projects such as professional baseball AI predictions and robot news article services. Even for him, the coin market is difficult to predict. He said, "The cryptocurrency market is hard to judge with human insight," and added, "Even so-called experts often have biases when analyzing, but I was motivated to take on this challenge because I thought AI analysis could provide more effective results."

However, limitations are clear. The accuracy rates of each AI vary widely. CEO Oh said, "There were mentors that collapsed during the recent Luna downturn," emphasizing, "No one should blindly trust anyone in the cryptocurrency market." The accuracy rate of AIs used for short-term investments exceeds 50%, but investors must check signals in real time. AIs aiming for high returns have accuracy rates hovering in the low 40% range. Therefore, investors ultimately need to review the information themselves and make comprehensive judgments.

Despite these limitations, CEO Oh stresses that Cosign will definitely help investors. The reason is 'transparency.' Lab2i discloses not only the algorithms used in AI analysis but also all results, even if returns and accuracy rates are negative. These figures are provided in a time series format, and an automatic report explaining why such results were derived is generated through 'explainable AI (XAI·eXplainable AI),' which also provides information to investors.

Transparency is the core of investor protection, according to CEO Oh. He explained, "Even people called experts sometimes gave investment advice with biases based on our analysis," and added, "We have also confirmed many fraudulent schemes that lure people with promises of guaranteed profits using underground information." He continued, "Cryptocurrency investors suffer losses because there are no healthy sources of information," and said, "While we cannot directly protect investors, by providing honest and transparent figures, we can help investors protect themselves and make rational choices."

CEO Oh’s goal is to grow Cosign into a platform where other developers’ AIs and experts with know-how can participate, compete, and analyze together. He said, "AI has the advantage of quickly analyzing large amounts of data, and humans have the edge in comprehensive judgment," adding, "If AI and humans compete and analyze together, we may discover investment methods that yield good results."