According to the IDC forecast report, the AI market size in 2020 is predicted to be about 157B$. Of course, this figure is the sum of the various industries related to AI: hardware, cloud services, and software. In addition, this report predicts that the annual average growth rate (CAGR) will reach about 17.1%, which will reach 300B$ by 2024. The following link is an article about this.
According to several sources, global smartphone sales in 2019 are predicted to be around 400B$-500B$. It is not without factors to improve CAGR such as 5G technology, but it seems clear that the growth rate is gradually slowing due to the increased penetration. In a few years, the size of the AI market is expected to catch up with smartphones, and it is clear that at least AI seems to be moving toward a situation where AI can be called a big market, not a niche market.
It's not just these rosy predictions. Although sales are clearly increasing, most service and software sales come from simple automation systems and rule-based chatbots rather than deep learning, and there are reports that the 90% of Chinese AI startups is not profitable. Deep Mind's deficit has exceeded 570M$, and services that generate external sales are not yet clear. In other words, although the market is growing, it has to be seen that it has not yet been performing in terms of ROI, and whether or not we can solve these problems will be our main competitiveness in the future. If the competitive indicator so far has been SOTA, I think that ROI will become more important in the future.