China now has plenty of AI startups yet only a handful is truly successful and people are beginning to wonder why. Lee Kai-Fu, former Microsoft and Google executive in China whom today has shifted to form a venture capital for AI startups in China shared an interesting viewpoint on the “tough challenges” for AI startups. Lee Kai-Fu shares that most AI startup clients in China are enterprises (the business terminology is “To B”: To Business). This means that Chinese AI startups try to apply AI to products for businesses. Then, businesses use these products to enhance their operations (such as AI that efficiently analyze or manage internal Big Data) or to develop other products for their own clients.
In the developing phase, these AI startups start by inquiring their clients to seek solutions. Then, they will generalize the solution so that their product can be applied to other companies in similar industries. Meanwhile, they will start collecting a one-time charge when solving clients’ problems. Then, they will develop a business model that continually collects fees and shares income from business clients.
The challenge that startups need to acknowledge is that, naturally, To B businesses will experience slower growth and more difficulties when compared to To C (To Consumer). A case study is that the CEO of these AI startups is often the sales and advertising representative of the company. This is because To B clients expect to hear only from the CEO. Also, the CEO is required to build trust regarding their AI competency and, when possible, the startup must provide an additional professional sales team. This aspect differs from To C clients where consumers quickly spread word of mouth right after the first buzz.
Chinese AI startups experience three “tough challenges”. One is that AI has become widespread and is not too difficult to learn. Programmers with strong basics in mathematics and computer engineering can further expand the AI skill. Therefore, a startup with a weak business model or one that can’t scale fast enough will easily be copied or surpassed by competitors.
Two, the strongest and fastest trend in China at the moment is the “AI + Cloud”. Tech giants in China such as Alibaba, Tencent, or Baidu are parading their cloud products to serve businesses clients. These products utilize AI for general requirements in business management.
As a result, businesses in China think that it is better to purchase cloud from tech giants as it is much cheaper than hiring or purchasing from startups. Moreover, when new functions or updates are required, they prefer to use the in-house AI team rather than hiring startups.
Lastly, AI application in business management or product improvement has rapidly developed and altered. As a result, companies are choosing to delay until the technology is ready and has reached a saturated point before investing. On the contrary, most startups need funding for experiments and product development right at the initial phase. Therefore, it is difficult for Chinese AI startups to find business clients who will support their idea since the beginning.
These are the tough challenges for Chinese AI startups according to Lee Kai-Fu’s analysis. Chinese AI startup shouldn’t think that as AI is trending, it will provide quick return and success. Actually, it is no easy task for AI startups to become successful. If one aims for success, they need to innovate a business model and product that is a solution for clients in the industrial sector at a reasonable price. If the solution is generalized or too expensive, it will be tough to compete with Chinese tech giants and their AI + Cloud.
Likewise, AI startup pioneers in Thailand shall understand the rapidly changing AI business landscape in a global aspect. In the future, players such as tech giants will move to AI + Cloud and the cost of AI in businesses will drastically decrease. Therefore, one must always acknowledge whether their business model and products are outstanding enough to drive the startup forward amid the competitive AI business battlefield.