In recent years, partly thanks to increased computational power and data generation, artificial intelligence (AI) has advanced significantly.
While we are currently some time away from a ‘general AI’ that can do everything that humans can and at the same time (i.e. multitasking), we can see that there are a few ways in which AI is better than human intelligence, particularly in areas involving speed, logic, scale (for example, analysing millions of lines of data in milliseconds, something that would take humans hours if not days to do), as well as prediction and detection.
Let’s take digital marketing as an example. Brands and retailers all over the world are looking for ways to more effectively build customer relationships, and AI is already doing a great job of identifying the right people to reach out to, determining what to say to them and the best time to say it.
According to a survey conducted by Appier and Forrester on AI adoption in APAC, the retail sector is further ahead than other industries in implementation of AI. The study found that 56% of retailers in Asia Pacific (APAC) have either implemented or are expanding AI initiatives.
The bottleneck is when it comes to generating creative content. Let’s look at marketing again. We know that AI can look at consumer data to uncover behaviour patterns and preferences, and then provide a recommendation about which images, colours, text and so on are most likely to appeal to a shopper.
After that, a human creative person is required to create an advertisement based on this information. However, we are starting to move towards a point where an AI system will be able to take this next step and generate content to the same standard as a human.
We are already seeing the first applications of this: chatbots that make it hard for customers to know if they’re speaking to a person or a machine; and simple news articles generated from financial data, such as earnings reports.
These capabilities are the result of advanced research in the field of AI and the development of new models and algorithms, and we can expect to see more of this type of content creation particularly in industries such as music, art and storytelling.
For the short-to-medium term, say five to 10 years or so, outputs from AI will still require human oversight or ‘gatekeepers’. This will be important to make sure that creative content is accurate; appeals to human sensibilities; and accounts for regional and cultural differences.
Although the most research tends to happen in English, the more AI models are trained on local or regional data (such as APAC, Southeast Asia, etc.) the more we will see creative outputs aligned with those regions.
Gatekeepers are also important to prevent issues around fake content such as news or videos. These challenges will arise as the technology develops, and Asian countries including China and Singapore are already taking measures to combat fake content.
However, we can also anticipate that the AI will become sophisticated enough to alert human end-users in cases when accuracy or legitimacy might be a concern.
When thinking about AI and its impact more generally in the coming years, as mentioned, we are unlikely to see the advent of AI systems that are completely equal to humans, and in fact it’s not what we need.
We will continue to see AI applied to very specific problems, with separate AI systems that do one or two things extremely well, such as uncovering consumer insights, diagnosing health conditions or managing logistics.
The best of these systems will be designed with the human end-user in mind, allowing AI and human intelligence to work alongside each other, allowing both to do what they’re best at.
