20-05-2025
Niftier Than Clippy, SAP Reimagines Omnipresent AI For Business
BERLIN, GERMANY - AUGUST 09: In this photo illustration hands are typing on a keyboard of a tablet ... More on August 09, 2024 in Berlin, Germany. (Photo Illustration by Thomas Imo/Photothek via Getty Images)
Clippy failed. Microsoft's ill-fated office assistant Clippy (officially named Clippit) largely failed because humans already react to computers in the same way that they do with humans. Variously described as a 'tragic misunderstanding and misinterpretation of anthropomorphism' by academics, Microsoft's miscalculation was to humanize an element of computing power that we already consider to be human, resulting in Clippy being regarded as an annoying interloper.
One quarter of a century later, enterprise artificial intelligence has to do a whole lot better than 'it looks like you're writing a letter' and start to directly engage in business workflows as it draws upon the agentic and generative AI functions of the modern age. Not only does enterprise AI have to be pitched at the right level of physical embodiment, it also needs to understand language, sentiment, context and intent in ways far beyond the doomed paperclip.
Although lessons have been learned, the hype-cycle driving AI (let's do AI, let's add retrieval augmented generation, let's embrace large language models… no, wait, let's specialize via small language models too, let's deliver non-deterministic agentic AI functions and so on) is surely not immune to creating technology for technology's sake. What we need is technology for the sake of business, technology for the sake of workplace well-being and sustainable technology for the sake of good.
German-American softwarehause SAP thinks it can draw upon its track record in enterprise resource planning technologies and business-centric software services to make AI functions both performant and practically useful. The company calls its copilot software Joule (as in the unit of energy, named after the eminent English physicist) and it is a generative AI service that works to streamline and enhance user interactions inside SAP business systems. The company wants Joule to be regarded as an omnipresent copilot that extends across a 'collaborative network of AI agents' that work across systems and lines of business. The company promises productivity gains up to a somewhat arbitrary sounding figure of 30 percent, but surely massive inconsistency in that measurement occurs across different industry verticals, different business models and differing organizational structures.
'SAP combines the world's most powerful suite of business applications with uniquely rich data and the latest AI innovations to create a virtuous circle of customer value. With the expansion of Joule, our partnerships with leading AI pioneers and advancements in SAP Business Data Cloud, we're making business AI real as we drive digital transformations that help customers thrive in an increasingly unpredictable world,' said SAP CEO Christian Klein, in a canned statement.
Perhaps addressing the exact type of service delivery miscalculations referenced above, SAP details its development of Joule, saying that it interacts with business applications, making work more conversational and personalized. Apparently eschewing any notion of an AI assistant's risk of getting in the way, SAP is instead going all-in and now unveiling 'an omnipresent Joule' that can accompany business users throughout their day, in and out of the SAP application universe, to find data, surface business real-time insights and streamline workflows.
Joule's new ubiquity includes an 'action bar' powered by SAP WalkMe that studies user behavior across applications, turning Joule into an always-on, proactive AI that anticipates users' needs before they arise. Aquired in 2024, WalkMe is a digital adoption technology designed to help user training and onboarding i.e. not onboarding to a new job per se (although digital adoption would very arguably help at this point) onboarding here refers to the process of getting users familiar with the format, functions and formulations of an enterprise software application so that it is used effectively.
This software's actions always adhere to SAP's strict ethical AI guidelines that anticipate how use of AI may also create economic, political and societal challenges, depending on how it is used and implemented. The company lists its rules of ethics based on UNESCO's 'Recommendation on the Ethics of Artificial Intelligence', which aligns with internationally recognized principles in this arena.
According to Klein and team, SAP is also working with partners to offer ecosystem of AI agents that can execute end-to-end processes across customer experience, supply chain management, spend management, finance and human capital management workplace functions. The company also announced an 'operating system for AI' development to help build, deploy and scale AI solutions. Known as SAP AI Foundation, this technology provides software application developers with tools to build and run custom AI solutions.
'SAP also announced new intelligent applications for SAP Business Data Cloud. Intelligent applications help businesses streamline routine work through a combination of standard business metrics, AI models and integrated planning features. The new People Intelligence application, for instance, transforms people and skills data from SAP SuccessFactors HCM Suite into workforce insights and AI-driven recommendations, helping leaders optimize team performance, support employee growth and maintain compliance,' said the company.
Despite claims, SAP is not necessarily the 'first AI operating system' in this arena. Microsoft Azure Sphere OS is a Linux-based OS that runs Azure Sphere microcontrollers; initially considered an internet of things service for intelligent devices, this technology has moved with the times towards the AI space and today works in close proximity to Microsoft's AI cloud services.
A technology that might need less explanation in this space is IBM Watson OS with pre-integrated software services and tools that work across machine learning (including deep learning) and cognitive computing workloads. Built by Canonical, Ubuntu AI is a data science OS designed to develop machine learning models with additional management tooling to help oversee security controls and more. Also in this arena, we find Google Fuchsia and Tesla's AI OS.
Created by Walturn, the personally-named Steve is an AI OS that uses shared AI memory and natural language processing. Its makers say that Steve is a user's 'gateway to AI-driven efficiency' as it helps streamline how businesses adopt and integrate AI. It ensures smarter workflows, better decision-making and enhanced productivity. If by now you're thinking this sounds like what SAP is saying, there's certainly some commonality across the marketplace here.
According to Walturn's Flavia Trotolo, Hashim Hayat and Daheem Hayat, 'Traditional operating systems were primarily designed to manage hardware resources and provide user interfaces for general-purpose computing. However, the rapid advancements in AI necessitated the development of specialized AI operating systems optimized for machine learning, deep learning and neural network computations. [An] AI OS incorporates capabilities such as real-time processing, distributed computing, edge AI integration, and enhanced security protocols to facilitate efficient AI-driven applications.'
The team further states that AI operating systems are 'particularly relevant' in domains that require adaptive learning and autonomous decision-making. These systems use technologies, including AI accelerators, cloud-based AI services and federated learning frameworks, to deliver high-performance computing experiences tailored to AI workloads.
Balancing measures aimed at SAP notwithstanding, what the company will clearly bring forward is huge experience of business data, a massive and expansive understanding of where data flows go up, down, left and right inside modern digital workflows… and a very bottom line-aware understanding of where the buck starts and stops. In the ever-accelerating world of AI hype, that foundation in financial fundamentals may prove to be the company's defining factor.
