Hendra Tan, CEO of 6estates said that from his experience, there are 3 key things that he can summarize from his past experiences and what he has been doing in the past years to make AI mainstream.
First, for AI applications, he used to row up numbers of AI projects with the budget of more than $30 Million, starting from the business scenario. Every department try to apply AI. The second is the user itself. There are some thoughts that the accuracy of AI won't be 100% to provide the user. This kind of mindset needs to be changed. Third, They need to redesign the business process involving human + AI to achieve a certain objective. The terminology used by people right now is human in the loop.
In the end of his explanation, he said that if we don't manage the expectacions, it will become PoC again, and never end up in actual usage.
Lie Heng, Director of Synnex Metrodata Indonesia explained that coming from the industry perspective, they are dealing with a lot of companies as well, especially enterprises and medium-small businesses. They have seen the implementation of AI used. For example, chart programs and zoom meeting applications that are already mainstream enough to be used in industry, have image processing, video processing, and data analytics capability. In the last 2 years, there has been a lot of acceleration in the adoption of AI. For then, 52 companies all over the world have adopted AI to deal with this crisis situation.
“Even in today’s event, many of the supporting companies like Prosa.ai, Konvergen.ai, and Catapa, that we have seen using AI. Seeing this status of AI, we can keep going ahead to accelerate it more than ever.” Stated Lie Heng in his explanation.
In the session, it was also explained that in the banking industry, we can see from the credit card fraud case. The bank has been monitoring the credit card company. They know just by seeing your regular charges, and you have it in a place that is not regular, it will get recognized. That type of AI has been running for quite some time.
In image and content recognition cases, YouTube, for years has been using image recognition to track uploaded videos that violated copyright.
In financial institutions, the use of chatbots is more common.
Now, bringing it to the mainstream, big players like Google and Amazon definitely have ingredients to build solutions. However, it is not that ready for smaller companies. That is why we need to build the ecosystem and partners first that will take those ingredients. If you have a model of license plate recognition, and you want to put it in your drive-thru restaurant so you can manage to know your regular customer when they come back and start to make their order earlier, you need to train those models, and it is different in every country. That is why it needs an ecosystem.