Technology or “tech” as we usually land up referring to it, gives prime position in the hierarchy of things to engineers. That is to be expected. But this has, over time, resulted in non-technical managers being participants in engineering processes that they can barely understand. That is not to say that a political science or arts graduate who went to business school, does not comprehend technology. Quite the opposite. The pervasiveness of devices and digital services embedded in such devices, has created a highly aware workforce(as indeed, consumer base). But being involved in the production and commercialization is quite another matter. Time, pressure to perform and differences in opinion, all contribute to a lot of friction. Engineers who build products are busy building products. Technical overseers(increasingly called product managers now) are concerned with timelines, quality, workforce management and releases. But management does not begin and end here. If we go back to an earlier generation of managers, we will see the 4 Ps-product, price, place and promotions. There is no reason for these to change. What has changed is the extent to which managers need to manage technology, because technology is pervasive and directly impactful. The answer is not necessarily moving an engineer into such a role(though an engineering graduate often does become a manager). The best person may or may not be technically qualified in a formal sense. The best person needs to manage technology well. That is different from engineering, where you work hands-on in “building tech” or using “tech” to create things. Technology management leverages the knowhow, experience and production lines inside a firm to maximize it’s corporate objectives as well as to create new opportunities to grow. That is still done through the management of product, price, place(distribution) and promotion(advertising and PR). Very well. But what has changed that we need to consider this a new discipline of management study-especially now? Well, the process of creation has moved from a set of deterministic steps to a series of connected, statistical processes. Artificial intelligence enables machines to co-create or participate in co-creation. They do this through pattern matching, a statistical process. In many areas of the economy, they fulfil predictive tasks such as inferencing, which again is about statistics and mathematics. This is not necessarily the same as classical industrial automation, though today automation is triggered by prompts delivered to a machine. In other words, automation is becoming closely intertwined with AI. It is worth keeping in mind that automation, like AI, is the end-result of code. Soon, the two will be indistinguishable. In many cases, that is already so. Now, while AI or early AI may be largely about pattern matching and predictive algorithms, the discipline has moved faster than the trade. Whether we reach artificial general intelligence eventually is a matter of debate. But we can see how the bringing together of different, disparate technologies with Generative AI is creating new classes of processes and end-outputs. It is clear by now that technology management will really be about navigating through this constantly changing landscape and fulfiling the management objectives of a company. Consider this. The ask is the fulfilment of managerial objectives when the fundamental technological bedrock of the company is increasingly based on chance. That same bedrock is also going to be prone to rapid change or rapid obsolescence. This puts an enormous burden on engineers and other technical staff. And this is where management of technology comes in. I propose that the true tasks of the new-age technology manager are:- To communicateTo clarify and simplifyTo design processes of uncertaintyTo design products that arise out of uncertainty To communicate To clarify and simplify To design processes of uncertainty To design products that arise out of uncertainty As the use of technology evolves, it’s role in mediating communications between people as well as between people and machines will become more important than it ever was. At the same time, the role of technology has to be communicated as a benefit to all publics of a corporate. These two will run along parallel rails. This is also where ethical designing will become a key KRA. Unlike in the case of more conventional IT, ethics will be a primary area of contest for businesses. Without paying due attention to it and allocating a significant amount of resources to it, companies may struggle to survive. The internal environment of a technology driven company is a hothouse of emotions, conflicting priorities and individual ambitions. This will sharpen as the technology of uncertainty and the uncertainty of technology grows. Steering priorities that are financial in appearance but deeper in intrinsic nature, will need a discipline and framework that embraces the turmoil. The above is where you come to the processes of uncertainty. What are these processes? These are not agile or waterfall, on one hand. They are not “first principles” rituals or “zero to one”. Agile and waterfall have real value but they were built for a well-ordered world where technology stayed static and it’s outcomes were predetermined. As for the rest, they represent sub-cultural flags borrowed from different parts of the world. The processes of uncertainty may be broken down into the following: Predict what is to comePrototype based on the predictionUse uncertain data because no one will gift you clean, structured dataUse the prototype as a strawman. Now go back to your data.Now build your real prototypes based on your real data and keep churning them. Extract small and actionable nuggets from the above every week, because this is a never-ending cycle. Predict what is to come Prototype based on the prediction Use uncertain data because no one will gift you clean, structured data Use the prototype as a strawman. Now go back to your data. Now build your real prototypes based on your real data and keep churning them. Extract small and actionable nuggets from the above every week, because this is a never-ending cycle. This, in turn, leads to products built out of uncertainty. What products? My belief is that product development-and we see this in case of LLMs and their associated end-user applications-will see far frequent pivots that will make releases almost redundant. You still need to keep the discipline of the product life-cycle, so that you can ensure that budgetary issues remain under control. The demand, increasingly, will be for building applications and solutions that can provide a look at what might happen tommorow or the day after. The challenge for engineers will be to deliver these and also ensure that their company is not left behind in efficiency, better data preparation or a “better” algorithm. If you just build an app that tells you-for example-your bank balance, it is just a typical app that informs you about how much money you have. Today it is likely to also let you transfer funds, pay bills etc. What product development needs to-and can-do is to give a predictive picture of savings growth(for instance) or savings scenarios(under different macroenvironmental circumstances) to the customer. We are now entering the zone of actively displaying a certain degree of randomness when it comes to people’s wealth. After all, we are not guaranteeing that growth. But we are showing possibilities. If the end-user wants to send the bank a message, he or she should be able to use natural language and ask fairly complex queries. We are now beginning to design via statistical engineering. Far deeper changes would take place in the back-office of the bank. The ability to go beyond the extant facts, while staying ethically and factually compliant, is something that is a big step forward for a financial institution when dealing with customers. A few weeks ago, I prototyped a recommendation engine for a motorcycle dealer. Now you will not use conventional catalogues, including digital ones.
