AI as a ‘Boon for Good Research’

• Artificial intelligence tools can complement the work of researchers, helping them identify complex patterns within data and suggesting methods to improve research outcomes.
• AI can accelerate traditionally time-consuming and expensive research steps, enabling a simpler, more efficient process.
• Researchers can use AI tools both to understand connections across data and to provide insights into human sentiment and its effects on companies.

Transcript

Soumitra Dutta: [00:15] AI represents today a collection of approaches and technologies that, in fact, are very much catered to a data-intensive world. So if you have a lot of data about situations, AI can be very good as a tool or technique to be able to analyze patterns in the data and to support research in that specific domain.

[00:37] It’s like, for example, the many cases of AI tools today helping doctors recognize medical symptoms, which even medical doctors—trained professionals—are unable to recognize because the data itself can be complicated. And sometimes, there are many differences to data structures that can actually enter into the picture.

[00:55] So, if you look at research, a lot of research is really very quantitative. It depends on the analysis of data. It depends on being able to understand patterns in data and find links across them.

AI can be very good as a tool or technique to be able to analyze patterns in the data and to support research in that specific domain.

[01:08] AI can be a huge benefit to researchers in helping understand these patterns, identify the patterns, and maybe find ways to improve the patterns. A good example is the research you see today in pharmaceuticals, where AI is being used to find combinations of molecules that might lead to good pharmaceutical outcomes.

[01:28] And the whole process of finding which molecule combines which other molecule is a very time-consuming, expensive process that would take pharma companies, typically a lot of money and a lot of time to go through the process fully, manually. Today, AI actually simplifies the process dramatically.

[01:44] The same thing is happening in finance research, economics research, and marketing research. I think it’s very important to understand that AI represents a kind of category, of a new category of tools and techniques for doing data-intensive research, which is very important.

AI can be a huge benefit to researchers in helping understand these patterns, identify the patterns, and maybe find ways to improve the patterns.

[02:01] AI can also be used in some cases for analyzing large volumes of text. Let’s say, for example, you want to analyze the sentiment of society toward a certain product or company, or for a company’s efforts in ESG.

[02:17] That kind of sentiment analysis, or looking at 'what 'people say and 'how' what people say affects the company’s performance, for example, could be a very good research question.

[02:27] So, analyzing these large amounts of natural language data, that is typically very difficult to analyze by traditional techniques, can be done by AI quite easily. So, I think AI can be a huge boon for good research in the future.