Where Generative AI hits paydirt
McKinsey and others credit AI with enabling automation to take off. It’s a questionable assumption except in one key area where, until now, success has eluded almost everybody.
Automating business activity requires knowing how the real world really works. It sounds obvious but it’s why much automation activity has so far been confined to non-core business activities. Business has not found a reliable, repeatable way to provide an accurate picture of how it works and thus how it can be changed and improved effectively. The enduring failure rate of change projects, be they new enterprise systems or consultant-led business transformations, remains at 75% primarily because successful change is reliant upon a good understanding of what a business currently does, and therefore what it has to change from and to in order to effect the change.
AI relies on the quality of its data. If we haven’t been able to supply automation with accurate data, Generative AI is not directly going to solve the problem. However, AI does hold the key to at least identifying the root cause of change failure.
For Generative AI to work properly it requires a model of the data it has to analyse. With ‘raw’ data, for example what customers buy, data models have been used for years. While this addresses an element of business, such as aiding customer service roles, it doesn’t help business operations, since to date there has been no way to model the activity-centric nature of business.
Consultants have been ‘mapping’ business processes for decades, but the tools they use have come from IT and diagramming the flow of data to help develop new software functions. This is very much a two dimensional view and not suited to the three dimensional nature of the real world of business. Recognition of this has been very slowly advancing through the recognition of the value of the Digital Twin, a software replica of the real-world. Slowly, because Digital Twin applications have thus far focused on physical assets, such as products and buildings, rather than the activity of the businesses that create them. Gartner talk about a Digital Twin of the Organisation, but have no description of what that is and instead point to the software tools that are currently failing to address the issue. Even if people see the Digital Twin as a model of the real world, no one has yet devised a model of business, which like a Swiss watch is a complex set of relationships between its various components that have to fit together in order for it to work.
Generative AI developers understand the importance of the model behind the AI. As AI pushes into business, past the superficial sales help scenario, the need for a model of how business works will become better recognised, leading to the eventual solution.
First must come the recognition of the distinction between superficial AI and the deeper-seated AI that can direct business to become increasingly more productive and thus more profitable. The potential for AI-led business improvement is an order of magnitude greater than the Just-in-Time revolution that has transformed manufacturing productivity. Imagine having AI tell you how to reduce the time to do things, while increasing the quality and decreasing the cost. Not once, like a consulting assignment, but every day. To do this it needs high quality data, i.e. what is really happening in a business, not just what someone thinks is happening. This requires a model of the business to make sense of the data and to generate advice that will fuel change and lead to increased productivity, not create chaos and bankruptcy from ‘fake advice’ that has been exemplified by the high-profile cases where ChatGPT and others have got it wrong.
Generating consistent real-world data to feed AI requires business systems to be automated – fully automated, not just the computerised functions of enterprise systems such as ERP that omit manual activity and the myriad of localised functions on spreadsheets and point solution Cloud-based apps. The need for automation is being spurred on by simplistic software products such as Microsoft’s Power Automate, which is really in the same league as their Access software that has been used successfully for years by non-IT people to address functions that ‘big’ IT doesn’t cater for.
Consequently, we see the slowly increasing pressure from Digital Twin, AI, and ‘lite’ automation software, making apparent the need for a real-world model of business.
The solution is a digital Workplace, a Digital Twin model in which employees can ‘see’ themselves, acknowledge that their work is accurately defined, and through which they engage with automated processes that tell them when there is work to be done. The model needs the capability to enable it to fully automate the most complex business environment.
Adding Generative AI to this environment is a game-changer. It interprets the data from automation and its model relationships to identify improvements in all aspects of a business process, in duration, quality, cost, and any other parameters that the AI can be trained to identify, analyse and present to management, resulting in a consistent flow of value to the bottom line.
AptumX Workplace is just such a solution.
To date, despite Workplace demonstrating spectacular results in increasing productivity, it has been constrained by the relationship IT has with business management, and their data-centric, rather than activity-centric view of the world. The software that IT currently uses to try to introduce automation typically fails to deliver on its promise, yet IT often pushes back against Digital Twin because it transfers their power to business managers and exposes a significant weakness - their lack of understanding of the business they are trying to support.
AptumX Workplace’s demonstrable ability to enable a change in business process in literally minutes is not always sufficient to have business executives override their IT decision-makers. However, we believe that being able to demonstrate the productivity and profitability- increasing power of AI-enabled automation will tip the scales.
Contact us to find out more about AptumX Workplace.