If you're not paying for it, you become the product.
While AI advancements follow one another at a breakneck pace, reflecting on the past two decades can shed light on a few questions and paradoxes that arise from the emergence of AI copilots in our everyday productivity tools.
Data and context ownership. From hardware to software, from software to cloud, from cloud to AI: over the past two decades, tech companies have gradually provided individuals and businesses with as many "layers" of productivity. Data and content were the material of the cloud era, which privacy and IP laws have struggled to regulate. The emergence of AI rather underscores the primacy of context and generative transformation, which now questions our conventional understanding of IP, as well as our data retention strategies. Suppose AI copilots brought by third parties can ultimately summarize all our thoughts based on our conversations. At a larger scale, such third-party AI may be able to "grasp" a company's overall context, which certainly holds much monetization value.
Tendency toward oligopoly. Antitrust regulators are still scrutinizing unresolved cases from the previous decade—see the ongoing Google trial—while new gray areas emerge. The race to develop ever more powerful AI models has led to massive investments in compute capabilities, which only the largest hyperscalers can afford. Once again, here is a significant risk of economic concentration, which may lead to hampering innovation or eroding ultimate productivity as a result of cost inflation.
Productivity. A recent study has revealed that the chaotic spread of overlapping productivity tools could erode up to 30% of working time. If AI introduces new features on top of existing ones, it might add to the digital fatigue that fragments our attention and cognitive abilities, pulling us further away from the essence of our work. The widespread adoption of AI in business certainly holds massive productivity gains. However, akin to the Solow Paradox in economics, there is a concern that a fraction of such productivity gains may be offset by issues of interoperability, resistance to change, ethical challenges, or even an increase in cost related to the maintenance and management of AI.
Cost. Because all of this comes at a cost, both economic and environmental. A cost today and a cost tomorrow, as the technological landscape matures and reaches equilibrium. So as our everyday productivity tools start integrating AI copilots at little to no additional cost, it is worth considering, again, this saying: if you're not paying for it, you become the product. There is no gift there.
Luckily enough, alongside proprietary solutions, the open-source AI ecosystem is moving fast and getting more robust every day. This gives hope for a future, healthy counterbalance to blackbox models that may bring about apparent productivity gains but ultimately at our own expense.
