After decades of development, AI technology companies and new algorithms emerge one after another, and the technological dividends of image recognition and speech recognition are released rapidly. However, in the past decade, it is mainly based on deep learning algorithms. The industry ecosystem is clearer, the homogenization of single technology is obvious, and the competitiveness is mainly reflected in the richness and specific details of the landing scene. When multi-mode capability combination and enabling other industries, the combination of AI technology and professional knowledge becomes a trend. This paper mainly summarizes three types of AI technology business models and four types of major industry participants. It is suggested to pay attention to the following aspects:
1) Revenue composition and recognition method: due to the rich use scenarios of AI, if it is an integrated government enterprise project, it is difficult for a single manufacturer to have complete ability. It is usually dominated by hardware sales, the proportion of software revenue is low, driving down the overall profitability, and the revenue progress is affected by the rhythm of customer project acceptance, reducing operation indicators and affecting cash flow;
2) AI enabled industry or industry nurturing AI: the traditional hardware company represented by Hikvision has ranked first in the market share of the security industry for eight consecutive years. It has deep insight into the needs of downstream customers and strong supply chain management ability. On this basis, the market goal of developing AI ability is more clear; Similar tiktok AI capability is mainly based on today’s headline App, jitter ecology and other business needs, and then technology spillovers; Alibaba’s strengths are also related to e-commerce business and commodity recommendation. Pure technology manufacturers need to have sufficient experience in market demand and customer use scenarios in order to form differentiated competitive advantages.
3) Products with high degree of standardization and large-scale replication can solve the difficulty of profitability: at present, AI Unicorn manufacturers receive more customized and privatized project orders, which can not form the advantage of large-scale replication, resulting in low per capita benefits. Taking iFLYTEK as an example, to C products are highly standardized, which can dilute manufacturing costs, R & D and sales expenses and improve profitability with the increase of market share, the shipment of learning machines and other products and the subscription of xuezhi.com.
4) Data is the top priority of AI development: there is no doubt that the three element algorithm of artificial intelligence, computing power and data are the most important. At present, the homogenization of algorithms is serious, the computing power is excessive, and the amount of data is large but the precision is insufficient. Big data, data warehouse and data analysis related companies may become investment opportunities. It is recommended to pay attention to Longtou minglue technology, a data processing company.
5) Personnel costs may decline or usher in an industry inflection point: the AI market has a large amount of water and fish. With the landing of various industry scenarios, the market has a clearer understanding of AI and a wider range of optional suppliers. At present, the business model of the AI industry is similar to that of consulting companies. The supply of talents will increase in the future, and the cost of personnel after the collapse of the first market bubble may lead to a rise in profit level and a structural change in the industry.
Artificial intelligence technology has not progressed as expected. Artificial intelligence industry is a technology intensive industry. If key technologies fail to achieve breakthroughs and relevant performance indicators fail to meet expectations in the future, it may have an adverse impact on the development of the industry
The landing progress and industrial application of artificial intelligence were less than expected. Companies in the industry rely on technology industrialization for profit. If artificial intelligence solutions fail to accurately meet the market demand, or the market growth is less than expected or even the scale is down, the revenue and profit of Companies in the industry may be lower than expected
Sino US trade friction. Sino US trade frictions may block the R & D and exchange of artificial intelligence technology and the supply of upstream artificial intelligence chips, thus affecting the business results of Companies in the industry
Industry competition intensifies. Companies in the industry are not only facing the competition of large comprehensive technology enterprises such as Google and Huawei, but also facing the challenges of many innovative enterprises in the vertical field of artificial intelligence. The intensification of industry competition may increase the cost of talents and reduce the profitability of potential projects, which will have an adverse impact on the company
Risks of macroeconomic fluctuations and changes in macroeconomic policies. Companies in the industry are greatly affected by the capital investment of government departments and have strong correlation with national industrial policies and macro-economy. If the above capital investment drops in the future due to macroeconomic slowdown, industrial policy adjustment and other factors, the overall market demand will be adversely affected