[Adobe Stock]
Despite waning hype, GenAI still seeing strong tailwinds
While hype around generative AI continues, there is much more to the field than hyperbole, according to Rohit Tandon, AI & Insights Practice Leader for Deloitte LLP. In a keynote at AI4, Tandon noted, “With Gen AI coming in, more people are thinking AI is real and not hype because they’re starting to see results.”
Deloitte isn’t alone in its optimism. Bloomberg Intelligence predicts that generative AI will be worth $1.3 trillion by 2032, with Big Tech companies in a competitive race to stay ahead with the technology. The momentum has enough inertia to drive significant long-term changes in the broader economy.
Tandon emphasized the profound impact AI will have: “Every dollar of GDP will be influenced by AI. I truly believe, just like everything you do today is influenced by IT, in the future, every dollar spent will, in some shape or form, be influenced by AI.”
Another factor driving adoption is the widespread use of generative AI at home. “The reality of today’s world is, if you’re used to something at home, how can you not give access to the same level of capability in the office?” Tandon pointed out.
Big benefits for some ambitious early adopters
Rohit Tandon
Tandon referenced Deloitte research that found 91% of organizations expect productivity to increase as a result of genAI tools. Nearly half, 48%, predicted a substantial transformation of their business in the next one to three years.
One Deloitte client, a $10 billion industrial manufacturer, expects to uncover more than $150 million in potential annual value from genAI technology. This includes $42 million in operations, $29 million in customer support, and $24.5 million in finance.
Tandon shared two compelling examples of genAI’s impact:
Accelerating time-to-market. “One of our clients, a software products company, creates a new product ready to launch, then has to wait three months for all the sales collateral to be ready, all the technical manuals to be ready, in 23 different languages,” he said. Deloitte is partnering with them to slash this cycle time from three months to three weeks. “I’m hoping three days, but let’s say three weeks right now,” Tandon added.
Boosting customer support. Another client dealing with technical manuals saw a dramatic improvement in response times. Previously, when a customer called with a problem, it required someone with over 16 years of experience to research the issue, taking up to a week. “Fast forward to today — two hours, someone with two years of experience is able to come back to them,” Tandon said, resulting in “a huge amount of customer satisfaction.”
Organizations at different adoption levels
As the generative AI landscape evolves, Tandon observed varying degrees of adoption across organizations: “What we’re also seeing is that organizations are at different stages of adoption. Some of them are still just exploring, some are implementing stuff, trying things out, and others are scaling.”
To illustrate this point, Tandon shared an anecdote about a conversation with “the CEO of a respected chain of hotels.” The chain opted for a more cautious approach, aiming to be a genAI fast-follower rather than a trailblazer. Tandon explained, “I agreed that they should not be the leader. There is no dire business need for them to go down this path, so they’re going to wait and watch, and they’re going to be fast followers, and that’s absolutely fine.”
However, Tandon cautioned that for many organizations, lacking a coherent genAI game plan or focusing solely on low-hanging fruit use cases could be risky in the long run.
Getting Out of the Shallow End of the GenAI Swimming Pool
Realizing the full potential of genAI demands scaling beyond mere experimentation and initial pilots. Tandon used a vivid analogy to drive this point home:
“One second, you want to swim in the swimming pool in your backyard, and then you think you’re ready to swim across the Pacific Ocean? No, it doesn’t work that way. You need a different set of capabilities, a different level of ability to scale, a different level of cost, to be able to scale up your AI programs.”
This scaling process, however, comes with its own set of challenges. Trust remains a significant hurdle, encompassing both confidence in the technology’s reliability and concerns about job displacement. Tandon warned:
“One bad incident triggers a hundred people to kind of stop wanting to go forward. One breach triggers a thousand folks who say, ‘I’m not trusting this technology. This technology is not ready.'”
To counter such risks, Tandon stressed the need for education to identify best practices, which may vary according to industry. He emphasized, “It’s about trust and confidence. So a lot of work needs to go ahead—awareness, education.” Many employers, especially in finance, tech, and pharma, have prohibited generative AI tools in the workplace due to perceived risks of data breaches, intellectual property theft, or regulatory non-compliance.
To drive adoption in a way that minimizes risk, Tandon highlighted the need for strong leadership, cross-organizational buy-in, and a modular approach when dealing with genAI, including Large Language Models (LLMs). Tandon also stressed the importance of breaking the “data gravity” mindset, where inertia constrains decision-making: “Don’t use that as an excuse to say, ‘Oh, we’ll make the one choice and then stick to it’.”
Augmenting rather than replacing humans
While some experts, including Geoffrey Hinton, whose work helped pave the way for genAI, warn of potential mass unemployment, Tandon offers a more optimistic view. He believes AI will spur new job types, similar to previous tech eras:
“Humans will not be replaced, but augmented: Humans using AI will replace humans not using AI. New skill sets will be needed, and new job opportunities will be created.”
Tandon urges adaptation, stating, “Please be prepared that you will need new skill sets, new capabilities, new kinds of jobs.”
He drew a parallel to past technological shifts:
“Remember when technology came out, word processors came out, and there was talk of typing pools and jobs being lost? Technology is going to take all jobs away? No way. It created way more jobs than existed before. Please be ready for the same.”
Tandon also underscored the need for workforce adaptation: “Until now, [many enterprise companies using genAI] been focusing on coding. Well, that won’t be enough for tomorrow. You’ve gotta get creative,” Tandon said. “You gotta get more collaborative. There are a bunch of new things that you will have to learn, a new skill set you have to learn for the new world.”
