Agentic AI: if you’re not using it, you’re already behind

Agentic AI is the next frontier of the revolution sweeping through all industries, not just publishing. Shrinking development horizons have meant that already it is being used widely. At the World Economic Forum in Davos, there were warnings that businesses must act now or be left behind.

Speaking in Davos on a special episode of Take on Tomorrow, a podcast by PwC, Matt Wood, PwC’s Global and US Commercial Technology and Innovation Officer, highlighted the urgency felt by attendees to harness the potential of artificial intelligence beyond mere automation. “A lot of folks attending feel an increased level of urgency to deliver on the opportunity of artificial intelligence,” he said. The discussions centred on applications that could change healthcare delivery, foster creativity and tackle complex engineering challenges.

Agentic AI distinguishes itself from traditional generative AI through its goal-oriented functionalities. As explained by Wood, agentic AI can process tasks in a more autonomous manner than its predecessors, such as completing complex arrangements like booking flights or even managing substantial financial portfolios. This technology not only aims for efficiency but also seeks to solve more abstract problems over extended periods, refining its strategy dynamically.

The conversation highlighted significant applications of agentic AI, particularly in software development. Wood claimed that these AI agents are currently performing at a level comparable to entry-level developers. “The AI will build your software application. It’ll write all the code, deploy it into your internal systems… and generate training material,” he noted. Over the next few years, the sophistication of agentic AI is expected to progress significantly, positioning them as invaluable assets within organisations adapting to technological transformations.

As such, there exists a clear demand for businesses to incorporate these systems into their operations. Wood asserted: “If you’re building software today and you’re not using these systems, you are already at a disadvantage to your nearest competitor.”

Furthermore, early adopters are witnessing enhanced customer satisfaction metrics. Utilising agentic systems in contact centres has resulted in quicker resolutions and higher customer satisfaction scores, demonstrating the operational benefits of these technologies.

However, with the rapid adoption of AI comes the critical question of trust. Wood emphasised the importance of ensuring that users have confidence in the AI systems deployed within their organisations. “You can spend an infinite amount of money … and if you are not also investing in the trust in that system, you will see zero return on that investment,” he said. Establishing transparency in AI operations is crucial. Providing users with insight into how decisions are made can significantly enhance trust in these systems.

Discussing the responsible use of AI, Wood articulated that responsibility encompasses a much broader scope than merely ethical applications. It necessitates stringent governance of data sources, ensuring data privacy and refining AI models for specific tasks. He stressed that responsibility should be ingrained in every layer of AI deployment, from data management to the application of results, necessitating a comprehensive understanding of all facets involved.

Looking ahead, industry experts are particularly enthused about upcoming advancements in reasoning models within AI, which promise the ability to tackle complex issues more effectively. These models allow AI to “think” through problems rather than merely respond, potentially revolutionising how interactions and problem-solving scenarios unfold in real-time contexts. Wood expressed excitement over the prospect of AI systems developing capabilities rivaling those of experts in various fields, notably in solving advanced mathematical problems.

Source: Noah Wire Services