Responsible Business Intelligence (RBI) stems from the three concepts of:
The responsible use of Business Intelligence (BI)
The responsible use of Artificial Intelligence (AI)
The application of BI and AI by Responsible Businesses
In my last newsletter, I looked at how can governments and society can use Responsible Business Intelligence to measure success by: how well business and economic growth benefits all regions and all stakeholders; and balancing productivity, environmental, and community outcomes - now and in the future.
The vast majority of businesses want to run in a responsible manner - driving growth in support of economic and societal benefits for all. Businesses want clear metrics to track this growth and understand these benefits, and they want the support of their shareholders and all other stakeholders; but shareholders and other stakeholders often present disparate and opaque expectations for the business. The business challenge is how to understand and reconcile these expectations and deliver progress and growth against them.
In a complex world - where business leaders are presented with potentially billions of relevant and often conflicting inputs - context, comparison and prioritisation of business risks and opportunities is critical. Responsible Business Intelligence is about empowering all business leaders with AI-driven analytics and contextual intelligence to manage these risks and opportunities.
In this newsletter I aim to also provide some context on what Responsible Business Intelligence is not about - misconceptions that can otherwise lead us down a different path.
10 Common Misconceptions of Responsible Business Intelligence
Responsible Business Intelligence:
Is not about Environmental and Social Governance at any cost: Business growth is critical for economic (GDP) growth which is critical for growing economic and societal benefits for all. RBI enables business leaders to balance stakeholders' expectations for growth with stakeholders' expectations for the environment and society.
Is not about telling Business Leaders what to think: Business leaders own their vision and aspirations for their business, what the mission and objectives are for the business, and what activities they will focus on to reach these objectives. RBI enables business leaders to map these activities (the "known world") against reliable public domain data, and other stakeholder datasets, to help them understand how they are performing and how they can adapt to win.
Is not about telling Stakeholders what to think: Society, and the broad church of stakeholders that a business may serve, does not subscribe to one way of thinking or one set of values. RBI enables business leaders to understand what different stakeholder groups are thinking - from shareholders to grass roots activists - to ensure business decisions can be made in the context of all stakeholder insight (the "unknown world") rather than the "eco chamber" that any one group of stakeholders or business leaders might generally operate in.
Is not about being a Black-box System: Trust is key for businesses, governments, stakeholders and society at large. Without trust, a business cannot thrive. RBI enables business leaders to build trust within their business and with their stakeholders by operating with objective, accountable and auditable intelligence. Businesses can not tell stakeholders what to think, but by providing the provenance for why a decision is made, business leaders can build trust and stand by their position in a transparent manner.
Is not about Perpetuating Bias and Misinformation: Concerns regarding AI perpetuating bias and misinformation are varied and valid: human bias in historical training data; algorithms inadvertently favouring certain outcomes - potentially due to lack of diversity in AI development teams; deepfakes and highly convincing synthetic media; and AI-powered amplification of false information. RBI enables business leaders to operate on vast, trusted, real time datasets where the AI is fine-tuned to only provide clear, objective outputs with reference to reliable sources.
Is not about Technological Determinism: RBI is not about letting technology drive decisions without considering the human and societal context. RBI enables business leaders to promote a balanced approach where technology serves the broader expectations of equity and justice including: privacy and copyright, cybersecurity, system resilience, regulation and human agency.
Is not about Resistance to Change: RBI is not about resisting technological progress through fear of job loss, lack of trust, lack of expertise, organisational and legacy system inertia or comfort with the status quo. RBI enables business leaders to deliver intelligence through responsible innovation and the adoption of new technologies in a manner that aligns with ethical standards and societal needs as well as business needs: clear leadership and vision; fail-fast phased implementations with pilot projects, case studies and success stories; AI skills development and continuous learning; and cross-functional and cross-expertise collaboration.
Is not about exclusion of Human Oversight: AI systems may make decisions that are technically correct but ethically questionable, lacking human moral judgment or accountability. They also need the trust and acceptance of human oversight to promote adoption. RBI enables business leaders to be the Human in the Loop (HITL), where AI decisions and responses are tested and reviewed before they are finalised. Business leaders can also provide feedback to RBI allowing for continuous improvement - and scenario planning for new and unforeseen situations - based on real-world inputs and human guidance.
Is not about Irresponsible Use of BI & AI: The vast majority of businesses want to run in a responsible manner - but not every business or actor fits into this bracket - there will always be some that wish to: manipulate public opinion, consumer behaviour, or stock markets; invade privacy; target vulnerabilities; discriminate and falsify; commit fraud; steal; extort; damage; and destroy. RBI enables business leaders to identify some of these challenges as they emerge; however, it is the combination of well-planned, regularly-tested technology and human systems and process, implemented diligently by the business, that will ultimately define their resilience.
Is not about extending LLMs at the expense of the Environment: According to Scott Stephenson a single trained Large Language Model (LLM) such as ChatGPT3 emits the same amount of carbon dioxide as a single trained (16 year old) human. Given there are only a few LLMs and there are billions of people, this shouldn't be a problem? Even though the number of LLMs is growing and the number of parameters in a an LLM is growing - 10 times as many parameters in ChatGPT4 as there are in ChatGPT3 - this only increases pre-training emissions equivalent to 10 trained humans per LLM.
Potentially, the real environmental concern is what happens when a large proportion of those billions of humans use the LLMs once trained.
Generating an image using a powerful AI model takes as much energy as fully charging your smartphone - effectively doubling the daily emissions of a smartphone using AI - resulting in 15 megatons of carbon dioxide globally over a year. Perhaps still small compared to the 63 megatons from bitcoin mining or the 350 megatons from datacenters overall - noting this is similar to the entire annual emissions of the UK. Nevertheless, Bill Gates said last month at the Breakthrough Energy Summit in London,
"Let’s not go overboard on this. Datacenters are, in the most extreme case, a 6% addition [in energy demand] but probably only 2% to 2.5%. The question is, will AI accelerate a more than 6% reduction? And the answer is: certainly.”
So, according to Gates, this challenge is one for AI to solve - through increased clean energy usage, and by making technology and electricity grids more efficient.
Comentarios