Where to start your AI Journey as an organisation?
Within the broad spectrum of AI one subtopic has come to the forefront in recent months, driven by the widespread use of ChatGPT and new product launches such as Microsoft Copilot: Generative AI (GenAI). For organisations, the question remains whether this will be a trend or an opportunity. At onepoint, we believe that Generative AI is set to be one of the most significant technological revolutions since the advent of computers, for the first time in the History, the gift of language is being bestowed upon AI. Organisations will be impacted at all levels, with a direct addition of value to the development of the organisation‘s business activities (humans will have much more time to devote to customer relations) and on the control of its resources (many administrative tasks will be automated and placed under human control, provided that these tasks are language-based). As a result, we are witnessing both a curious “wait and see” trend among organisations, which are experimenting as they can in an environment that changes daily, and a frenzied race to add value, akin to the gold rush of the 19th century. We advise our clients to be part of this fundamental movement, and, at a minimum, to anticipate what their business will look like in a few years (or even months!), rather than risk losing their competitive advantage. Having said that, how does an organisation go about tackling transforming themselves into a Generative AI proficient business?
Know thyself: the need for organisational evaluation to understand the effort required to implement AI
First, let’s deconstruct the myth of AI: this technology is a continuity of tools and processes that an organisation might have already implemented. For instance, if the usage of Cloud is already democratised within the organisation it will be a great accelerator for any AI project as it is one of the foundations to deploy it. Another example would be Data Management: if an organisation has already invested in defining relationships between information, knowing its data assets, regularly assessing its data quality, and other data governance practices, this will also be an accelerator for AI projects.
This is why the ancient adage “know thyself” has never been more apt. By evaluating its maturity level in key areas, an organisation can have an initial estimation of the financial, technical, and governance efforts required to implement Generative AI models. Understanding these efforts will also give an organisation high-level visibility on investments it needs to make, complimenting other analysis and allowing the organisation to produce an ROI evaluation.
Prepare the organisation: the 3 core challenges to consider before beginning
For an organisation, embarking on the generative AI journey is not just about embracing a new technology. The expected upheaval in the coming years (or even months, given the rapid pace) is not about acquiring just another solution but more about recognising and tackling complex challenges:
- Financial Obstacles: Implementing generative AI solutions is not cheap. The uncertainty and the lack of prior experience makes early adopters pioneers. Securing high-performance computing infrastructure and obtaining the right kind of data involves significant financial considerations. For instance, an organisation may face the daunting upfront cost of several million dollars for the necessary computing infrastructure, before even considering the ongoing maintenance and data acquisition expenses.
- Technical Constraints: Money isn’t everything. Is the organisation equipped with the necessary hardware, software, and data infrastructure to bolster Generative AI efforts? What legacy technical challenges does the organisation need to manage? For example, an organisation may want to implement a PDF summarizer, but if the documents aren’t OCR processed, it won’t work!
- Human Capital: Cutting-edge systems still heavily depend on human expertise. Does the team possess the nuanced expertise in AI, machine learning, and data science to unlock the full potential of generative AI? Are business unit aligned staff able to explain and document the rules that the LLMs will have to follow?
Evaluate your maturity: Leverage the AI Readiness Assessment Tool to understand what is required for a successful AI journey
To help address these challenges, onepoint has developed an AI Readiness Assessment Tool. The tool is designed not to uncover the value that generative AI can bring to an organisation, but rather to map out the conditions and prerequisites necessary to enable an organisation to assess the value of uses cases and enable their implementation. Divided in 5 axes detailed below, this AI Readiness Assessment tool aims to define the effort and capabilities required to both design and implement Generative AI opportunities.
- Every project should start with a Strategy and Vision and Generative AI projects are no exception to this. The rapid innovation pace of this technology makes attempts to have a static multi-year vision for generative AI usage within an organisation futile. More than defining an idea(l) state, the objective of this axis is to understand the organisation’s current position when facing this opportunity and what levers are available for it pull:: what is the level of sponsorship? What is the level of risk aversion? How much is the organisation ready to invest? Tomorrow, why not imagine that an insurer is only present in the pocket of the insured, or that a travel operator focuses its customer relationship channel solely on a WhatsApp conversation?
- The second axis is Human Capital and Organisation. An orchestra is only as good as its conductor and musicians level and synchronisation. Similarly, the potential of AI is realised when steered by a harmonious team of experts, a continuous learning ethos, interdisciplinary collaboration, and a network of external partnerships. These skills don’t come out of the blue: is there a Generative AI recruitment plan? Is there a training plan? What about an upskilling plan to capitalise on your existing talents? Coming back to the orchestra metaphor, it is only going to sound good with harmonised scores: are your processes documented and synchronised?
- It is impossible to separate AI from the Technology landscape it will operate in. Within this CIO-oriented axis, enablers or blockers to the AI journey can be identified, such as, the usage of cloud services or the level of interoperability between information systems. A legacy on-premises and “closed” system is more challenging to implement an AI solution with when compared to a cloud-based infrastructure. In addition, what is the level of control over the organisation IT infrastructure and projects? Some companies have federated control models whereas other may have centralised approval processes which can take up to several months to review and approve technology changes.
- The fourth axis is Data, and as for everything, what is provided as inputs will determine the quality of outputs. Poor data inputs will lead to poor answers, or even worse, wrong answers. It isn’t just about volume, but a symphony of diversity, privacy, ethics, and meticulous annotation. Is the data identified? Documented? Digitalized (OCRed)? Is there a mature data governance process in place?
- Last but far from being least, the fifth axis deals with the Environment. By gathering Compliance and Societal considerations, this aims at building strong Generative AI ethical convictions at the organisation scale. With great power comes great responsibility: Generative AI is a tool that can have direct interactions with an organisation’s clients and internal talent. It involves crafting comprehensive guidelines for the ethical application of Generative AI, advocating transparency, staying updated with ever-evolving AI-specific regulations, and consistently guarding against biases.
To conclude, the allure of Generative AI is indeed vast, and its adoption is potentially transformative, yet it hinges crucially on an organisation’s willingness to embark on a profound introspective journey. In an era where technological revolutions emerge with promise and pace, it is vital for organisations to pause and meticulously invest in understanding their unique strengths and vulnerabilities. As we navigate through these exhilarating times where every strategic bet can lead to significant leaps forward or setbacks, a thorough self-assessment is not just advisable; it’s imperative. Knowing oneself deeply is the essential first step before stepping decisively into the arena of these enthralling innovations. By delving deep into the fabric of their capabilities and readiness, organisations can not only brace for a future sculpted by AI but can also seize the reins to shape that future, thriving amidst the whirlwind of change and redefining the boundaries of what’s achievable.
