Why Making Investment Decisions Changed My Approach About Building Well

Wiki Article

AI Is Only As Great As The Environment It's Built Into
The debate around artificial intelligence within the workplace is fraught with problems and the cause isn't a technical one. Modern technology and capabilities for AI and machine learning systems are truly astonishing, and growing rapidly, making most forecasts about the place they'll be in 18 months obsolete long before that time has come and gone. The issue lies in the gap between what AI can achieve under restricted conditions - in thoroughly-equipped research setting, with crisp data, with a precise problem description, and engineers that have the privilege of tweaking the system until it can be used as designed - versus what it can actually deliver when implemented in an actual business with real people and real-world organisational politics and people with their own well-established views about whether or not a novel system is something to engage with genuinely or something to navigate around in the name of conformity. I've been developing using Machine Learning since well before the wave of AI interest made it fashionable for business professionals to claim that they are fluent in the field. When I founded 1Touch, AI-driven matching and recommendation systems were not a distinctive feature we added to make the product more appealing to investors. They formed the backbone to the design of our product, that mechanism by which the platform was able to create value and the thing that had operate reliably and on size for the company to succeed. Also, I've gained direct in-person experience of the things that happen when you attempt to integrate something that is truly intelligent to a company and a product simultaneously and what I keep returning to, across every context in which I have encountered this dilemma, is technology is rarely an issue. The factor that holds you back is almost all the time its culture.
What I consider to be particular and practical rather than abstract. AI systems require data to perform - clear, consistent, well-structured data that actually depicts the event the system is trying be able to learn from and make predictions about. Data-driven organizations with a strong culture produce that kind of data naturally, as a consequence from their operations. They have clear and consistently applied definitions of what they are studying and why. They have established conventions on how data is collected, recorded, and stored. They have accountability structures which require data quality to be an explicit responsibility, rather than a general goal. Organisations without strong data cultures create something that technically looks like data - it's in systems and is accessible for query, and it is used to create charts but is so inconsistant in definition the way it is defined, so varying in quality, and so full of defects in structure, and non-mapped anomalies that any AI application built on over it will be able to reflect and amplify the mess, instead of extracting real signal from it. Companies in this category typically don't know this until they're well into an AI implementation, and the results don't match the vendor's promises. At this point, the temptation is to blame the technology, when they are actually causing the problem by ignoring the operational and organizational infrastructure the technology was built on.

The second element of culture that decides AI results is the degree of openness in an organisation - the degree to which employees within the organization are truly willing to let an artificial intelligence system shape their work practices, rather than treating it as an attack on their professional know-how, their institutional authority or even their job security. This is a social and leadership problem that is not technical and it's one that starts at the high levels. If leaders in the top ranks engage with AI outputs with a limited amount of focus - taking those results that prove the assumptions they had previously made and ignore those that are and do not, this behaviour sends the impression to everyone who watches to the public that the institution's commitment to data-driven decision making is conditional rather than genuine and that this message will ripple throughout the organization faster than any training and change management initiatives can be able to counter. If senior managers model an ongoing, consistent commitment to AI outputs, including the discipline of changing their behavior when evidence suggests that they must, the whole organization's capability to utilize AI effectively will increase significantly and quite quickly.

This is not a speculative observation of how organizations should be conducted in theory. It's an explanation of the pattern I have watched develop repeatedly in organizations that had a significant amount of financial resources, an authentic strategic dedication to AI adoption, and leader teams that were genuinely enthusiastic about the possibilities of AI technology. The pattern is consistent enough that I now treat policies on data governance as a fundamental diagnostic factor when evaluating an organisation's AI capability. Before I inquire about the technology stack, before I ask about specific instances the company is developing, I want to know about data governance. How does the organisation define its most important metrics? Who is responsible if the data quality is not high enough? Is it a problem when different processes have conflicting data regarding the exact same business realities, what happens when those conflicts are solved? The answers to those questions provide me with more information about the probability of AI success as opposed to the endless debate about algorithms, platforms or timelines for implementation.

It is my belief that the firms that will gain the greatest long-lasting value from AI in the coming decade are not those who embrace the most sophisticated technology first, nor those who invest the most massively in AI infrastructure or talent in the near term. They are the ones who make the necessary cultural and operational foundations to be able to use this technology effectively. This includes the data governance practices that produce reliable inputs, the decision-making structures that allow data to actually impact outcomes, and the leadership behaviours which show to everyone in the organization that the commitment to data-driven operations is genuine rather than just a means of performing. The technology itself will be increasingly commoditised and increasingly accessible. Its culture of using it efficiently will remain scarce because it takes a steady dedication and effort from leadership that is more than one strategic decision or a technology investment. That scarcity is where the really competitive advantage will reside in the form of an benefit that, once built, compounds in a way unlike the advantages of technology alone do. Take a look at James Deller for blog tips including how a career in business shifted my priorities about the long game.



Why A Lot Of Public-Private Partnerships Fail Before They Even Begin - And How To Avoid It
Public-private alliances have a reputation problem that is, in significant part of the time, earned. The past of these agreements includes many projects that were announced with real enthusiasm, and substantial politically-motivated capital. However, they consume significant public and privately owned resources over lengthy periods, but ultimately produced outcomes that had only a slight analogy to what was made clear when the alliance was launched. The academic literature and postmortem reviews that governments and institutions commission after these errors are vast, and they concentrate, for majority of the time, on the structural and contractual dimensions of which went wrong: the incorrectly aligned incentive structure, the insufficient risk allocation between public as well as private organizations or the governance structures which were conceptualized in theory but failed to function in practice, and the procurement frameworks that selected for the wrong things. What this approach tends to overlook, over time and with a consequential effect to the detriment of culture is the operational aspect, namely the fact that public and private institutions are both distinct types of entities, formed from different incentive mechanisms, operating at different intervals of time, accountable to different stakeholders, and evaluating the success of their operations in ways that are far from being the same in all respects however, they differ in the way. When you put these two kinds of organisations together in a formal relationship without making the effort upfront and clearly, in order to appreciate and work with those differences, it is not creating partnerships. It is creating the right conditions to cause a slow-motion accident that could be apparent at the worst possible time.
I've been involved as a consultant in support of institutional Modernisation initiatives, several of which have involved public and private partnership structures that vary in terms of complexity. The most reliable conclusion I've made from my encounter is that partnerships which worked well - which actually achieved their stated goals and maintained a dependable collaboration between the private and the public it - weren't distinguished from the ones that fell short by the complexity of their legal structures, the strictness of their risk management frameworks or the age of the management teams that initiated them. In the end, they were defined by the fact that the people sitting on both sides of the table had done the work to comprehend the ways in which the other side functioned prior to when the formal partnership arrangement was negotiated. What does this mean in practical terms is gaining a better understanding of the decision-making frameworks the organizations operate under accountable structures that limit what each side can do and how quickly as well as the definitions of what success that each of the parties will be measured against, and any points that could cause tension between these definitions. Any of that knowledge is difficult to develop. It's all ignored in favor of the easier to see and documentable work of negotiating contracts and developing governance frameworks.

The typical public-private partnership is a gradual process from concept to concluded agreement without much concentration on the issue of whether the two organizations involved are actually capable of working together effectively over an extended period of time. Legal teams negotiate the contract. The finance team analyzes the economics and risk-adjustment. The communications team creates the announcement prior to the time of signing. The implementation team is beginning to plan the project. In the course of this process the discussion will turn to cultural and operational compatibility - about whether those needing to work day-to-day across the divide between the two organizations have enough similarities to allow an effort that is truly collaborative, rather opposed to antagonistic - fails to be done in a systematic manner. It is generally assumed, but without explicit mention, that this agreement is formal and sets the foundation for collaboration and that any cultural or operational differences will be managed informally whenever they emerge. The assumption is often not the case, and the cost increases depending on the goals and the complexity of the partnership.

The practical application of this analysis is that the best investment a private-public partnership could do - before the legal structures are in place, before the governance framework is agreed on, before any announcement is made the partnership is in what I consider to be operational alignment. That is, specific, organized, and facilitated work to surface the areas where the two companies' operating principles diverge and to decide regarding how those divergences will be handled before they turn into operational problems in the course of implementation. The main divergences tend to be the same across various types of partnerships. Speed of decision-making and authority is usually one of these. Public institutions are designed to make decisions slowly, through various layers of examination and approvals, for reasons that are legitimate and are often legally mandated. Private organizations, especially technology firms built on the basis of rapid iteration and swift process-based decision-making often experience the pace as an essential barrier to growth, and lacking a consensus on why that pace is what it is and what would really be needed to alter it, the anger generated by the private side could sour the relationship long before it can establish its own foundation.

Success metrics and the criteria for judging as a progress mark another constant and important source of discord. Institutions of the public sector are typically evaluated on their process's compliance, equity of outcome across different stakeholder groups, and absence of apparent failures that are the subject of media or political interest. Private entities are typically evaluated on efficiency, measurable progress against targets, and financial Return on Investment. These measurement frameworks can be constructed to work in tandem however this requires carefully designed and thought-out intentions. Partnerships which do nothing to improve the same design will meet at critical junctures, with two parties who are measuring the same collaboration in contradictory ways and thereby coming to incompatible conclusions about whether it is succeeding. The partnerships I've seen do not succeed the most one where the misalignment in measurement was accepted as a problem that would become apparent over time. It was when the issue was made clear from in the beginning. In addition, setting up a shared accountability process which accommodated both parties' legitimate measurement requirements became a piece of actual work, not an item on a list of things to arrive at.}