Three AI predictions for 2024

AI has evolved beyond the confines of tech businesses to become part of our everyday lives. And in 2023, generative AI (gen AI) exploded onto the scene, opening doors to exciting opportunities across every industry.

From spotting fraud in banking to personalizing treatment in healthcare and even creating new flavors in the food industry, gen AI is creating innovative new ways for people to get work done. Undoubtedly, gen AI will play a key role for businesses in 2024.

Here are three ways I predict gen-AI capabilities will unfold in the year ahead:

Foundational models (FMs), the essential building blocks of gen-AI applications, will continue to get larger and more powerful. But the larger the model, the higher the training costs – which could make many proofs of concept unaffordable.

Fortunately, the open-source large language model (LLM) community is addressing this challenge. Some are building gen-AI pilot projects using tiny LLMs. Customizing these smaller language models for a specific business or industry is easily achievable and offers a more cost-effective solution. Even the creators of powerful closed-source models are beginning to dish out custom generative pre-trained transformers (GPT).

All these developments are pushing businesses toward custom enterprise FMs. While the trade-offs of performance and privacy will dictate the choice of model, the debate around open-source versus closed-source LLMs will only heat up moving into 2024.

Gen AI can handle more than just text. Multimodal tools can understand and extract insights from multiple data types, like images, video, audio, and more. These capabilities help engineers build AI agents fine-tuned to perform a string of tasks, while LLM-powered agents can enable back offices to run 24/7.

For example, ChatGPT's voice feature eliminates the need for tedious typing. Users simply voice their instructions while the LLM application fetches a summary, creates charts, plans the fiscal year, and more. And with gen AI handling these time-consuming tasks, people can take on more strategic and innovative work.

You can see similar examples in the travel industry, where multimodal agents actively function as personal trip planners. The user only has to say "plan my holiday" to an app and give the LLM-powered agent access to their calendar. The agent will skim text, scan images, and generate a travel itinerary. It's a massive time-saver and productivity boost for travel agents, while customers enjoy quick and highly personalized recommendations.

However, as AI applications become more sophisticated, monitoring AI agents is crucial. Using agent operations (AgentOps), with analytics and debugging capabilities, is vital to mitigate risks like errors and biases in planning, task decomposition, reflection, and execution processes. Human oversight is also necessary to address these issues promptly. Likewise, embedding responsible gen-AI principles throughout the process is critical.

Customer experience (CX) is one area where gen AI is gaining traction. Gen AI can answer questions more efficiently, resolve problems on first contact, communicate clearly, and even make recommendations. In the year ahead, we'll see more businesses finding new ways to measure and reimagine customer experiences – all thanks to gen AI.

The CX industry will also witness a rise in AI-powered contact center solutions. Tools such as contact center AI (CCAI) and contact center as a service (CCaaS) are already doing the rounds. Here, gen AI empowers human agents with continuous support during customer calls and chats by identifying purchase intent and providing real-time, step-by-step assistance. The technology can also identify customer motivations and inclinations, helping contact center managers learn about customer interactions to improve the service they provide.

It's something we're already doing at Genpact. We helped a global media brand incorporate personalized gen-AI recommendations into its customer service chats – resulting in better experiences for agents and customers.

In 2024, AI will continue to enhance the way we work, increasing the need to develop new data science skills at a breakneck speed. Gen AI will also cut down go-to-market time for enterprises, which calls for agile leadership and ongoing change management capabilities.

The key to success will be to fundamentally redesign processes from the ground up, delivering value that extends beyond single productivity gains but transforms the fabric of business operations. By blending the power of AI with human ingenuity, enterprise leaders will usher in a new era of business innovation.