Way back in 1997, IBM’s Deep Blue supercomputer faced off against world chess champion Garry Kasparov. In a stunning showcase of early artificial intelligence, Deep Blue outsmarted the pro to win the six-game match. In 2011, IBM’s Watson AI won the US quiz show Jeopardy! Five years later in 2016, DeepMind’s AlphaGo won a match of Go (a game several magnitudes more complex than chess) against Lee Sedol, another world champion. Fast forward to 2021 and in a landmark moment for AI, DeepMind’s AlphaFold solved a long-standing problem in bioscience and predicted protein structures.
For decades, we have seen AI consistently push the boundaries of what’s possible. And yet, the level of excitement around AI today is unprecedented. Everyone from the world’s best scientists to college students in dorm rooms is attempting to integrate AI into everyday life and work. What’s driving this sudden surge in interest? Perhaps, prefixing “AI” with the word “generative” holds some answers.
Unlike its predecessors that crunched data and identified patterns, generative AI takes a bolder leap – it creates; assembling pixels into realistic images, composing music with nothing but data, and writing marketing copy capable of actually resonating with audiences. It’s no wonder a recent PWC report estimated that AI could pump $15.7 Trillion into the global economy by 2030, and generative AI is a big part of that.
Part hype or all substance?
If you’ve been wondering what’s driving the hype, join the club. As an Engineering First organization, we’d be remiss if we didn’t try to find answers. Analyzing the buzz, two things stand out: people are captivated by how generative AI mimics human cognition, and how it is much more approachable than any traditional AI system of the past.
Here are three key reasons why we think generative AI is more substance than hype:
- Unlocking potential
Stop us if you’ve heard this before: Generative AI is here to help humans, not replace them. It’s a statement that’s made for many-a-clichéd blog titles – but only because it’s so true. A 2017 McKinsey study stated that while automation could displace some jobs, it will also create new ones. Generative AI supports this argument – especially in the era of knowledge workers looking to do more with less. A legal professional can leverage generative AI to quickly compare legal documents, freeing up time for strategic analysis and client interaction. Businesses looking to increase their focus on the big picture will need generative AI to handle their repetitive tasks, unlocking thousands of hours of human potential.
- Unending evolution
OpenAI announced the game-changing GPT-3 LLM in November of 2022. At which point you could ask it to write a poem about cats and it would string together words into a largely nonsensical thought. Just a year later, GPT-4 could write that poem in the style of Edgar Allan Poe in a perfectly executed iambic pentameter. Thanks to its inherent learning and attention capabilities, generative AI is growing at an exponential rate. Google’s Gemini surpassed its not-very-old prior models in a remarkably short period and Anthropic’s latest AI models are outperforming almost everything in benchmarks. The message for enterprises is clear: staying ahead of the curve in generative AI is crucial for maintaining a competitive edge.
- Unprecedented productivity
Generative AI’s assistive capabilities and its rapid evolution add up to a sum that can be hard to comprehend. “Traditional” AI systems like Deep Blue or AlphaFold tackled specific tasks with incredible results. However, generative AI has popularized the view that AI can tackle complete workflows and act more as a collaborative “assistant” than a point tool. Let’s say you’re a wealth management advisor tasked with preparing for your next meeting with an important client. With the right AI system, you could analyze your client’s personal and portfolio information, collect relevant documents from a variety of repositories, and assemble the information that you need to share during the meeting, with the option to adjust tonality based on notes you might have made on the client’s personality. This ability to “daisy chain” tasks unlocks a level of productivity companies could only dream of back in 2020.
Virtusa’s commitment to generative AI
Generative AI is here to stay. And at Virtusa, we are firmly committed to this transformative technology’s development and implementation. Since its inflection point in 2023, we’ve been paying attention to how the technology is progressing and its long-term implications. A key realization was that generative AI while creating hype, has also provided a shot in the arm for broader AI adoption. But to ensure successful AI adoption, companies must first identify the right processes where AI can add value, and then break them down into workflows that can be augmented by AI.
This presents challenges for companies that lack the internal expertise to navigate the fast-evolving landscape and complexities of AI solutions and adoption. To help ease the process, we have built a generative AI center of excellence responsible for staying on top of the technology’s rapid evolution. This includes following and studying generative AI developments from CSPs like AWS, Google Cloud, and Azure, as well as diving into the world of open-source models that can deliver significant cost and customization upsides. Our center of excellence also focuses on broader AI adoption combining generative AI and traditional AI to build compound AI systems.
To broaden our exploration of generative AI within the firm, we have also established a generative AI guild that is open to any employee who wants to learn and experiment with this exciting new technology. The end goal is to create a powerhouse that will help our clients uncover and implement AI strategies that work for their specific needs.
An Engineering First approach
Generative AI adoption will be at its most effective when the technology is woven into business operations, instead of being patched onto individual parts. Virtusa’s Engineering First approach is designed to unlock this effectiveness by rejecting a point-problem approach and instead looking at processes end to end, to identify exactly what needs AI assistance, and analyzing how this impacts the steps in the process before or after.
Our steadfast focus on effective data utilization also ensures that clients derive maximum value from their investments.
But building practical solutions that work in the real world requires going beyond creating proofs-of-concept and focusing on quality, safety, and production-grade certification. Every solution we create is a result of iterative development, incessant testing, and engineering solutions on the go, as we identify problems. The result is a solution that’s high on accuracy, and thus more usable and reliable for users in real-world scenarios.
Ultimately, Virtusa aims to identify and mitigate the challenges of a nascent technology that holds a ton of promise for individuals, enterprises, and society at large. While it's hard to identify all potential challenges that a technology like generative AI can pose, having an engineering mindset enables us to spot, understand, and address problems along the way.
Stay tuned to learn more about how Virtusa is gearing up to unleash generative AI’s potential.