The Journey of Generative AI

When OpenAI announced ChatGPT4 to the world last November, the mood shifted from awe to apprehension rather quickly. Hardly a year is out and some green shoots of enthusiasm seem to be appearing now. In the interregnum, businesses responded in a polarized fashion, with some going on an exploratory overdrive and others simply banning its use in their workplaces.

So, what prompted these diverse reactions to something that merely involves models that study common patterns and arrangements across massive data sets using computing techniques that we’ve always known as deep-learning. The almost instant results could either be convincing or full of gaps based on what one is using it for. So, where’s the mystery?

What’s making the word so polarizing?

For starters, it is worth mentioning that everyone is experimenting. Some sectors have achieved success and others haven’t. At a more granular level, some enterprises have succeeded in using the models in certain industry verticals while others haven’t found their mojo yet. So, we have healthcare, finance, digital marketing, and education giving a thumbs up at this point. 

At the other end of this spectrum, over 75% of global businesses have implemented or are considering bans on ChatGPT and other generative AI applications at their workplaces. A research by Blackberry found that 61% of the companies consider these bans long-term on account of risks to data security, privacy and corporate reputation, says a report. 

However, consulting companies think otherwise

Amidst this polarized scenario, an interesting facet is emerging where top consulting companies including The Big Four accounting firms are quite kosher with generative AI. A report published in AnalyticsMag quotes Sachin Arora, partner and head at Lighthouse (Data, AI, and Analytics), KPMG India to suggest that a ChatGPT like framework is being used internally. 

In fact, the company is spending $2 billion for an alliance with Microsoft, which we assume is part of its overall $5 billion spend that it had earmarked back in 2019 for AI-based technology. Now, if you thought that’s an isolated incident, here’s some more. PwC has set aside $1 billion towards advancing its generative AI involvement in the US over the next three years. 

Once again it is Microsoft that’s at the forefront. The company aims to automate certain elements of its tax, audit and consulting functions via OpenAI models, all of which is aimed at developing AI and GenAI applications that improve efficiencies, reduce costs, and cut down time. Of course, what we are unaware of now is how much of human capital will be eliminated. 

Are the enterprises ready to absorb GenAI?

Meanwhile, another large consulting company EY has integrated AI into its operations around payroll queries as well as tax laws where a ChatGPT like interface provides instant answers to queries. The company has since reported better accuracy and increased efficiencies. Last, but not the least, even Deloitte launched a GenAI practice in India back in June. 

Romal Shetty, chief executive officer, Deloitte South Asia, was confident that more and more enterprises would jump on to the GenAI bandwagon in the months ahead which is why the company decided to set up a practice in the region. It aims to guide clients in understanding the risks, limitations and ethical considerations of generative AI. 

Finally, consultancy giant McKinsey also struck a deal with AI startup Cohere to provide AI solutions to its enterprise customers. This is the first such partnership with a large language model provider for Mckinsey, which joined other consulting companies as the frenzy over ChatGPT kicked off across the world. 

And what’s the future that consultants see?

So, where is all this leading to? Arora believes that the early reluctance amongst companies is going away as they stemmed from data security and privacy worries, ethical concerns and the regulatory challenges. He believes that companies such as KPMG have a role to play in making AI more accessible and user-friendly. 

Pre-built AI platforms facilitated experimentation, leading to successful generative AI projects. Innovations in data management addressed data privacy apprehensions, with privacy-preserving AI techniques ensuring secure implementation. The establishment of ethical guidelines and standards enabled responsible AI usage and mitigated ethical dilemmas tied to AI-generated content, says the article. 

In the short-term, all of this could lead to more instances of GenAI integration with traditional AI and analytics resulting in a substantial boost to employee productivity. And as the journey continues, it could result in quicker rollout of new products and services based on the disruption of conventional information brokering practices, Arora concludes.