In an interview with Sadika Sayyad, Prakash Pimpale discussed the advanced AI and NLP technologies catering to global industries. He stated that global industries have progressed from an offline world to a digitized world over the last many years and now the way ahead is the automation of as many activities as possible. This automation involves imparting intelligence to machines through AI and NLP.
Communication has become the primary medium for modern business as a result of the advancements in the digital age. Global companies are implementing complex and advanced AI and NLP technologies as communication takes center stage in the commercial environment.
Prakash Pimpale : Artificial Intelligence (AI) and Natural Language Processing (NLP) have been topics of interest in R&D labs for a long. In recent years, both technologies have found their grounds in various industries through different use cases and in turn, huge resources and newer problems have become available for the R&D labs to solve. This has accelerated the development of newer and better AI and NLP technologies like GPT, LaMDA, IBM Watson, Siri, Alexa, etc.
The scale and speed at which these technologies are getting adopted is evident from their current and estimated market size. According to some reputed sources, the global artificial intelligence (AI) market was valued at nearly USD 60 billion in 2021 and is estimated to reach around USD 1000 billion by 2030.
In my opinion, this adoption is very natural. The global industries have progressed from an offline world to a digitised world over the last many years and now the way ahead is automation of as many activities as possible. This automation involves imparting intelligence to the machines through AI. And huge part of this automation will need to deal with natural languages like English, Hindi, Marathi, Russian, Chinese, etc. as language is an important aspect of our life. So, NLP being a subfield of AI has a lot to contribute to this automation.
With this adoption we will also need to consider and act on its impact on availability of jobs, data privacy issues and other societal issues among others.
Prakash Pimpale : Many industries including Healthcare, Software services, Education, Automotive, Financial services and Agriculture among others are using NLP in numerous ways. They are using it to improve/automate their operations or better their customer services. Some common use cases which have been identified and implemented can be enumerated as:
- Text Classification for email filtering, ticket/customer complaint classification, document classification, Sentiment Analysis, etc.
- Named Entity Recognition for huge number of tasks where the information is to be extracted from the natural language text. For example, extracting information about companies and their performances from huge financial reports available as text documents.
- Question Answering and Chatbots are popular in L1 level support across businesses.
- Translation across languages by the computers has been gaining wide acceptance.
- Natural Language Generation/Content Creation is an exciting use case that will see a very speedy adoption in coming years with the availability of efficient Large Langue Models (LLMs).
- Summarization is being used to summarise the large financial reports and documents.
- Knowledge Graph like applications are being used in drug discovery to speed up the process, in customer services to understand users better or in business to take better decisions.
Prakash Pimpale : The speed at which industries are adopting AI and NLP has resulted in a huge demand for skilled resources in these technologies. The newly trained resources are too fresh to understand the business and contribute effectively to solving a burning business problem. But this will get resolved over the next couple of years.
Another challenge that occurs to me is that of over-expectations from these technologies or a lack of understanding of their exact capabilities. The C-suites should be equipped to understand where exactly the AI should be fitted in the business.
Over the next couple of years, companies must solve the challenges of data privacy, multilingualism in the data, ways to augment the errors of AI systems, and cost of the AI adoption for certain problems which may not seem feasible.
Among others, the challenges of Bias in the AI systems, ethical concerns, and other societal issues need our attention to keep the pace of adoption growing.
Sadika: How can AI and NLP help businesses to enhance omnichannel engagement and connect with customers?
Prakash Pimpale: Language being an important part of human communication, NLP technologies can play a significant role here along with other AI technologies.
- Businesses can make use of AI-powered chatbots to provide customer support and answer FAQs through different channels such as websites, e-mails, messaging apps, and social media.
- Predictive Analytics based on user data can be used to make personalized recommendations to users and serve them better.
- Sentiment Analysis should be used to monitor customer satisfaction on various social media platforms and during the conversation with the business.
Prakash Pimpale: I think the coming years are very exciting for the AI and NLP. The end of the last year’s chatGPT release has given a glimpse of what will be offered ahead. The web search is going to get changed significantly. There will lot of activity over the cloud on easing development and deployment of the AI and NLP solutions. AutoNLP seems to be another trend that will accelerate the adoption of NLP technologies by augmenting the lack of skilled manpower.
The development of advanced Machine Learning technologies is also going to be an important catalyst for the progress in these areas. The need for large data sets for solving certain problems will reduce. Obviously, we aim to achieve learning like a kid who learns to avoid walls or any similar obstacles by getting hit by them twice or thrice and not million times!
In the spotlight
Prakash Pimpale works as a Joint Director with C-DAC. Centre for Development of Advanced Computing (C-DAC) is the premier R&D organization of the Ministry of Electronics and Information Technology (MeitY) for carrying out R&D in IT, Electronics and associated areas.
He has been involved in various projects in the areas of Machine Translation, Natural Language Processing, Machine Learning and Data Science.
