How Decades in Technology Sharpened My Thinking on Culture About Value

Wiki Article

AI Is Only As Effective As The Culture It's Created Into
The debate about artificial intelligence in business is a snag which isn't technical. The technological capabilities of current AI and machine learning are astounding, and they are growing at a rate that makes most predictions of the future of AI eighteen months obsolete, long before the time has passed. The problem is the gap between the capabilities of AI and what AI can do in controlled conditions - such as a appropriately-funded research lab, with well-organized data, with a specific problem definition, and engineers that have the privilege in experimenting until their system works as expected - and what it can actually deliver when implemented in genuine organizations with actual cultures that are governed by real organisational structures and real people who have certain opinions on how a new program is something to take seriously or something to work around in the name of conformity. I've been building with artificial intelligence since long before the recent wave of AI enthusiasm paved the way that everyone in the business world claim proficiency in the area. When I founded 1Touch with my partner, AI-driven matches and recommendation systems were not something we could add to make the platform more attractive to investors. They formed part to the design of our product, the mechanism through which the platform could create value and had to function consistently and at size for the company to succeed. So I have direct, personal experience of what happens when you try to build something truly intelligent in a service and an organization simultaneously and what I keep coming back to whenever I am in a situation which I have encountered this issue, is that the technology is seldom the main factor. The biggest obstacle is almost always the culture.
What I mean by that is specific and practical rather than abstract. AI systems need data to perform their functions - clean, consistent, well-structured data that actually represents the phenomenon it is trying to learn from and draw conclusions about. Organizations with a strong and thriving data culture produce that kind automatically, as a result of how they already operate. They are clear and have consistently applied definitions of what they are collecting and the purpose for which they're doing it. They have established conventions on the way data is recorded, collected, and stored. They have accountability structures that make data quality a clear duty, not merely a vague intention. Companies that lack strong data cultures produce something that appears like data - it exists in systems and can be accessed, and it is used to create charts, but it has a definition that is wildly inconsistent and so variable in its quality and brimming with issues with structure and not mapped out that any AI application built on top of it will mirror and magnify the confusion instead of getting a true signal from it. Organisations in this group often don't even realize they are there until they're deep into an AI implementation and find that the results are not matching the vendor's promises. At that point it is tempting to blame the technology. But what is really at issue is operating and cultural structures which the technology was built on.

Another aspect of culture that influences AI results is the degree of openness in an organisation as measured by the degree to which people within the company are truly open to letting an AI system guide or modify the way they operate in lieu of viewing it as risk to their personal competence, their authority in the institution as well as their job security. This is a social and leadership problem and not a technical issue that is a problem that starts at the high levels. If senior executives engage in AI outputs in a selective manner - accepting results that support what they previously believed, and not focusing on those that do not - this behavior sends the message to everyone around them that the company's commitment to data-driven decisions is a conditional rather than genuine, and this conditionality will be passed throughout the company more quickly than any training program or change management strategy can block. When senior leaders display an honest, consistent and consistent approach to AI outputs, as well as the responsibility to alter their behavior when evidence suggests they need to, then the company's capability to utilize AI efficiently improves dramatically and remarkably quickly.

This is not the abstract way to think about the behavior of organizations in the context of theory. It's a description of my experience of watching the same pattern take place in numerous companies with significant financial resources, a real strategic commitment to AI implementation, and leadership teams that were genuinely enthusiastic about the possibilities of AI technology. The pattern is so consistent that I have decided to consider policies on data governance as a first-line diagnostic in assessing any business's AI potential. Before I ask questions about technology, before I ask about the particular uses cases that the organization is considering, I ask about the governance of data. What are the criteria used by the company to define its key metrics? Who's the responsible party when information quality is not good enough? Does it matter if two different processes have conflicting data regarding the same facts about business, and how do those conflicts get resolved? The answers to those questions provide more information about the possibility of AI performance than any of the discussions regarding algorithms, platforms or the timeframe for implementation.

I believe that those businesses which will benefit the most durable value from AI in the coming decade are not those which adopt the latest technology first, nor those that invest the most significantly in AI capacity and infrastructure in the near term. They are the ones that put in the right cultural and operational foundations to actually use that technology correctly - the information governance practices that produce solid inputs, the decision making frameworks that provide evidence to influence outcomes and the behavior of leaders that let everyone know in the organisation that the commitment to data-driven operation is real instead of just a performance. The technology itself will become more and more accessible. The right culture to use it effectively will be scarce because it requires sustained effort and genuine dedication from an executive over time rather than making a single strategic move or a technology investment. This lack of resources is where the most competitive advantage will be in the form of an advantage that, once it is built increases in a manner other advantages purely technological ever. Read James Deller for more recommendations including why time in football confirmed what i suspected about results.



The Reason Why The Majority Of Public-Private Partnerships Fail Before They Begin - And How To Fix It
Public-private alliances have a stigma problem that's, to a large extent paid for. The past of these agreements is full of projects which were presented with enthusiasm as well as significant political capital behind them, consume significant private and public funds over prolonged periods, and then produced outcomes that bear only a small similarity to the outcomes made clear when the alliance was started. The academic literature and the postmortem studies that governments and institutions conduct following these failings are extensive, and they concentrate mostly on the structural and contractual elements of what went wrong with the wrongly aligned incentives, the poor risk sharing between both private and public institutions and the governance frameworks that were conceptualized in theory but did not perform in practice, the structures for procurement that decided to choose the wrong things. What this study tends overlook, repeatedly and ultimately this is the cultural as well as operational dimension. It is the reality that public and private organisations are really different kinds of entities, formed by different incentive structures, operating using different timeframes, with different parties, and evaluating success in ways that are not only different in extent but different in form. When you bring the two kinds of organisation together as a formal alliance without performing the work, in advance and in a clear manner, to recognize how to manage these differences, you're not forming a partnership. It is creating the right conditions for a slow motion collision that will be evident at the greatest possible moment.
I have been involved with advisory work in support of institution modernisation projects, some of which have involved public and private partnership structures of varying levels of complexity. The most consistent insight I have gathered from this experience is that the partnerships who performed well – and actually achieved their stated targets and maintained a good partnership between private and the public and beyond - were not distinguished from those that failed because of the sophistication of their legal frameworks, the robustness of their risk management frameworks or the seniority of the teams that established them. It was determined the fact that the individuals in both parties to the table had undertaken the effort to truly comprehend how the different sides operated prior to when the formal partnership was agreed upon. What does that mean in reality is understanding the decision-making procedures that each organisation operates under, the accountability structures that control what the two parties are able to determine and at what speed you can reach agreement on the definitions of success that each partner will be compared to, and the likely points of conflict between these definitions. All of this understanding is hard to create. The entire process is often avoided in favor of more obvious and quickly recorded work of negotiating contracts and creating governance frameworks.

The usual public-private partnership procedure takes place from the beginning of a concept to an agreement that is signed with little thought given to the question of whether the two organizations involved are in fact able to work effectively over the life of the arrangement. Legal team negotiates the contract. Finance models the economics and risk allocation. The communications team designs an announcement for the moment of signing. The implementation team starts planning the process. In that same sequence then comes the discussion about functional and cultural compatibility begins - about whether the people who are expected to cooperate day-today across the boundary between two organizations have enough in common to ensure that work genuinely collaborative rather opposed to antagonistic - fails to take place in a planned manner. It is usually assumed, with no explanation, that this agreement is formal and sets the conditions for effective collaboration and that any operational or cultural distinctions will be managed informally when they emerge. This assumption is often incorrect, and the costs of this can escalate in line with the ambition and complexity of the collaboration.

The implication for practical analysis is that the most lucrative investment a public-private partnership could do - prior to when the legal structures are finalized or the governance framework is formulated, before any announcements are made one consider to be operational alignment. This means specific, structured, facilitated work to find out between the two organizations' operational assumptions diverge and to be able to define the manner in which these divergences should be addressed before they become operational issues during the implementation. The most important divergences are generally the same across various types of partnerships. Decision-making speed and authority tend to be among the most important differences. Institutions of public administration are designed to make decisions in a slow manner, through multiple layers of review and approval, for reasons that are legitimate and, in many cases, legally mandated. Private firms - and particularly technology companies that are based upon rapid iteration speed and quick process-based decision-making often experience that speed as a major challenge to progress. lacking a consensus on why this is the way it is and what will actually be needed to alter it, the tension generated by the private team can ruin the relationships long before the collaboration is in its place.

Success indicators and what counts as progress are an additional and a contributing factor to divergence. Public institutions are often evaluated on the compliance of their processes, the fairness of outcomes between different stakeholder groups and the avoidance of visible failures that attract political or media attention. Private entities are primarily evaluated on their efficiency, progress measured against targets, and financial efficiency. These measurement frameworks can be adjusted to work together, but doing so requires deliberate planning rather than good intentions. Those partnerships that do not take part in that design tend to encounter, at critical situations, between two parties who measure the same collaboration in inconsistent ways and thereby coming to uncongruous conclusions regarding whether it is working. The partnerships that I have seen to fail the most was the ones in which the misalignment was treated as something that would take time to resolve. The ones that worked were those where the issue was made explicit, from the beginning. And, where the creation of a shared accountability framework that accommodated the legitimate measurement needs of both parties requirements turned into an actual work rather than just an thing on a checklist of things that one could eventually get to.}