26 OPINION 01732 759725 Randy Grohs, Chief Architect at Nintex, explains how to keep humans onside when integrating AI into business processes businesses in the next five to ten years, so, as Gartner says, it should be a core principle not a project. More haste, less speed In the excitement to implement AI, many teams rush to utilise the next iteration to avoid falling behind. But rushing mindlessly to layer new versions of AI on top of traditional automation may see some fall short. A lack of understanding when it comes to implementing AI can have significant repercussions for both employees and management. Critical mistakes and unintended consequences are probable without a comprehensive grasp of the technology, including its strengths, weaknesses and potential impact on tasks. Furthermore, incorrect data analysis becomes a looming concern when AI systems are not adequately programmed or supervised. This can result in erroneous decisions and predictions. Security is a critical aspect of any AI implementation but often gets overlooked in the rush to adopt this technology. AI systems, like any digital counterparts, are vulnerable to cyber threats. Hastily implemented AI may lack the necessary security measures, exposing businesses to potential cyber-attacks and data breaches. Inadequate testing of AI systems is another risk associated with rapid implementation. Failing to conduct thorough testing can lead to system errors or failures once the AI is integrated into the workflow, disrupting operations and efficiency. Ethical concerns loom large in the context of swift AI implementation. AI can potentially infringe privacy and engage in behaviour that raises ethical dilemmas. When implemented hastily, there may not be adequate time to establish safeguards and regulations. Given the powerful capabilities of AI, it becomes essential to carefully steer and, in some instances, restrict the AI engine to ensure it operates within policy and maintains its intended character and focus. Inaccuracy presents another challenge, particularly in generative AI, which can demonstrate considerable creativity. While creativity can be desirable, it can also lead to unpredictable outcomes if not carefully directed. Is governance the key to success? When RPA first burst onto the scene, companies introduced a legion of new bots, often with inadequate governance. Suddenly, companies didn’t have visibility into how much automation was being used. Implementing disparate AI solutions across the enterprise without a way of coordinating and governing them invites the same mistakes again. When it comes to incorporating AI, being methodical, doing proper analysis, fine-tuning and properly restricting how AI is used, performing adequate testing and keeping humans at the centre are all key ingredients to getting things right from the outset, without alienating the workforce. Fine tuning the AI recipe It is crucial to get the AI recipe right before attempting to scale. Automation projects suffer when there is a lack of coordination between stakeholders or collective automation goals are not properly understood. AI is heavily used in some forms of automation, such as intelligent document processing (IDP), which uses computer vision AI to perform OCR and document recognition, then automates the processing of the resulting digitised information. Despite the practice having been around for many years, the overall process involves ‘humans in the loop’ to handle discrepancies and exceptions because the technology still isn’t 100% accurate. In the rapidly evolving landscape of AI adoption, one central truth is clear: the paramount importance of keeping humans at the forefront of every initiative. As the allure of AI's transformative power beckons businesses forward, we must remember that the true measure of success lies in our ability to harness its potential while safeguarding the wellbeing of our teams. Technology should empower, not replace, humanity. www.nintex.com The demand for artificial intelligence (AI) in business is rising, with organisations recognising its potential to drive innovation, efficiency and growth. According to a recent study by Deloitte (The State of AI in the Enterprise, Fifth Edition), over 50% of companies plan to incorporate AI and automation technologies into their operations in 2023. As the opportunities for AI continue to surge, it is crucial for businesses to reflect on the valuable lessons learned from previous automation implementations. Rushing headlong into AI adoption without learning from past automation mistakes can inadvertently hinder its effectiveness. In 2019, EY found that 30% to 50% of initial robotic process automation (RPA) projects failed, often because companies fell into the trap of implementing the technology first and thinking about the problem it was meant to solve later, resulting in diminished productivity gains and preventing organisation-wide buy-in. The importance of getting AI right Considering the speed at which AI is developing and its potential impact on how we work, it is vital for business leaders to avoid making the same mistakes again. Simply deploying AI at a functional level is not enough. To unlock its full potential, enterprises must aim to progress towards true scale and scope, transcending initial implementations to achieve transformative, enterprise-wide integration. And to accomplish this, organisations must proceed with vigilance and a deep understanding of previous automation journeys. AI adoption's primary focus should be creating a new level of business value. By assimilating the lessons learned from previous automation endeavours, business leaders can avoid repeating mistakes and make more informed decisions maximising the chances of their AI initiatives delivering on their promises and contributing to a genuinely transformative business environment. The AI landscape may make or break AI integration: balancing technology and people Randy Grohs
RkJQdWJsaXNoZXIy NDUxNDM=