The world has been enthralled with OpenAI’s ChatGPT since its launch last November, as it showed how even non-technical users could easily use generative AI to create new content using existing materials. According to data from analytics firm Similarweb, an average of about 13 million unique visitors used ChatGPT per day in January, more than double the levels of December 2022.
From everyday users to organisations alike, many are quickly adopting the generative AI tool in their daily functions. For instance, a Snapchat creator invented her own CarynAI chatbot “virtual girlfriend” based on ChatGPT, while a Seattle desk worker lost 26 pounds after turning ChatGPT into his fitness trainer.
Beyond personal use, generative AI has various business applications too. According to tech research and consulting firm Gartner, venture capital firms invested over US$1.7 billion ($2.28 billion) in generative AI solutions over the last three years, with AI-enabled drug discovery and AI software coding receiving the most funding. The generative AI market is expected to grow, with market research firm MarketsandMarkets anticipating reaching US$51.8 billion by 2028.
AI trading assistant
Tiger Brokers is one the latest organisations to jump onto the bandwagon. The brokerage firm has developed its own text-generating AI chatbot called TigerGPT, based on OpenAI’s ChatGPT, as an investment assistance tool to be launched on its Tiger Trade platform soon.
With TigerGPT, users can get a consolidated view of a listed company — from its price-to-earnings ratio, net present value, financial earnings, and stock movement insights to recent affairs insights — to make informed investment decisions in a few seconds. The information is pulled from the brokerage’s content library and its expansive access to paid sources on the web.
Additionally, TigerGPT can help users improve their financial knowledge by explaining more advanced topics like portfolio diversification and risk management in an easy-to-understand, human-like dialogue manner.
Jacques Li, Tiger Brokers’ head of global communications, says this development resulted from the AI revolution that took the world by storm in late 2022 as a practical alternative to human creators. “We started to keep a close eye on how AI is going from November last year,” says Li, who shares that TigerGPT was launched internally this January.
The chatbot has been offered to selected users of the Tiger Brokers platform who have opted in to test the programme since April 11. Users are not required to have a brokerage account with Tiger Brokers, but they must sign up on the trading platform.
TigerGPT rides on the wave of financial institutions that have launched versions of text-generating AI chatbots, such as Bloomberg and Morgan Stanley. However, TigerGPT exists to prepare investors before they invest, and it does not provide investment strategies or solicit stocks. Li believes this is TigerGPT’s unique advantage, in that the firm is committed to “raise the [knowledge] efficiency [of its users] before making investments”.
To train TigerGPT’s AI model, the product team at Tiger Brokers feeds it with a hoard of topics, questions that they think users might have, and the feedback they receive from users. Over 3,000 users across various markets — Singapore, Indonesia, Malaysia, Philippines, Australia, New Zealand and Hong Kong — have trialled TigerGPT so far. The chatbot uses users’ questions and responses to help train its responsiveness and accuracy of information provided.
However, Li says that TigerGPT does not collect or retain its users’ personal information. It merely picks up user feedback about its features, and responses cannot be tracked back to any single individual.
Data privacy concerns
The AI revolution has come under fire for concerns about managing individuals’ data privacy of late. Large language models such as ChatGPT are trained on vast swathes of Internet data and the information users enter into the system, often involving private details.
According to CNN, some companies, including JPMorgan Chase, have clamped down on employees’ use of ChatGPT due to compliance concerns related to employees’ use of third-party software. But Li says that it “does not make sense” to keep any personal information of its users as investment information is “general and generic”. “[Instead, TigerGPT serves as] a guide to offer timely and accurate market data.”
The brokerage industry, he adds, is tightly regulated and requires firms to comply strictly. It is, therefore, not in the company’s interest to breach any privacy concerns with its new AI chatbot. He also asserts that the firm takes a hard stance on AI systems and technologies, which should be subjected to rigorous safety evaluations and adopted properly, ethically and legitimately.
Li says that the firm is actively interacting with regulators in different markets on how AI could best benefit its users within the regulatory framework. “Compliance is the bottom line of how this industry thrives, which is quite important [to us].”