ChatGPT and the likes: cause to celebrate or call ‘mayday’?

Ever since the news of Microsoft investing a whopping $10 billion in OpenAI broke, the business and tech worlds haven’t been the same. The word echoing through publications and social media was ChatGPT. A riveting step up from the search engines of the world, this tool not only crawls the web to suggest results but also strings them into coherent answers. Less than an hour of use will tell you the enormity of subjects and data it commands, amplifying the apprehensions we’ve always had about Artificial Intelligence- how will this help or hinder our workforce?

While one to stand out, ChatGPT is far from being the only tool that uses advanced AI technology to solve problems pertaining to coding, creative writing, mathematics, technical writing, graphic designing, data analysis and more. Among the other prominent AI technology being developed are TensorFlow by Google, PyTorch by Facebook, GPT-3 (ChatGPT) by OpenAI; Watson by IBM and more.

ChatGPT, short for “Generative Pre-trained Transformer”, is OpenAI’s potent language generation model and is a useful tool for a variety of applications, such as chatbots, text summarization, and content generation as it can produce human-like language responses to a given cue. Responding to users in a conversational manner by using a natural language processing (NLP) system, technologies like this can come in handy across domains and job roles.

• Customer Support: AI-driven chatbots & IVRs can provide 24/7 customer support, helping to answer customer queries quickly and effectively. With the ability to fetch information from the databases and convey it to the user, AI-driven chatbots and IVRs can answer frequently asked questions and resolve common issues, such as billing and account management.
• Sales and Marketing: Chatbots can assist in generating leads, qualifying prospects, and closing deals by guiding customers through the purchase process. AI can be used to provide personalized recommendations and offers to customers, which can increase customer engagement and sales.
• Education & Healthcare: AI-driven tools can be used to generate educational content, such as summaries, outlines, and study guides, based on a given text. Chatbots can help in diagnosing health problems and offer personalized medical advice based on the report.
• Travel & Entertainment: AI-driven tools can provide personalized recommendations and suggestions for travel destinations, activities, and accommodations based on the user’s preferences and budget. This can also apply to customized movies, music, and other recommendation services based on their usage.
• Business Intelligence: AI-driven chatbots can enable real-time insights into market trends, customer behavior, and more. ChatGPT can assist with business intelligence by providing natural language generation capabilities that can turn data into human-readable insights and reports. It can also assist with data analysis by providing natural language understanding capabilities that can process unstructured data and extract relevant information.

What it means for HR and talent management

A successful recruitment today is no easy feat. While skill gaps continue to create major hiring roadblocks, recruiters and hiring managers often find themselves knee-deep in processes, paperwork and repetitive tasks that waste valuable time and resources. From writing job descriptions to handling recurring and repetitive tasks, AI tools like ChatGPT may be the assistants a modern-day HR manager needs.

With their prompt responses and follow-ups, AI tools can enhance the candidate experience and lower the dropout rate. Besides, on the backend, they can assist recruiters to understand the role better, send outreach emails to candidates, assess candidates, manage salary benchmarking data and so on. Thus, solving the challenges of speed, quality and proactive engagement in the HR space.

However, as the world is all excited and ardent about these AI technologies, there are potential risks and downsides that one should be aware of. After all, it is just a tool in its preliminary stage and we should keep our expectations in check. We must be cognizant of the possibility that ChatGPT may sound reasonable even if its output is wrong. The model may generate text that is biased or discriminatory if it is trained on data that contains biases. Consequently, it could generate text that is misleading or false if it is not well-trained or if it is given incorrect information. There can be privacy concerns as the model may process and store personal data, and more importantly, such models may be used for malicious purposes, such as creating fake news, impersonating real people online or spreading misinformation which are all alarming scenarios.

Conclusion

While they are great for automating rule-based, repetitive, mundane tasks, these tools lack what technology till the foreseeable future will continue to miss- experience, instinct, and empathy, to name a few. GPT technology is best used to augment tasks rather than replace them. If coupled right with human intelligence, it holds the potential to make our tasks simpler, and time-efficient, while leaving enough room for intellectually stimulating work that provides high value.

(The article was first published in CXOtoday.)

Varun Sachdeva

SVP and APAC Head, NLB Services