• We argue that a cautious “wait and see” approach—motivated by gen AI’s flaws, such as hallucinations—is potentially dangerous.
• Nontechie employees can use gen AI without support from experts.
• What’s more, gen AI is increasingly being embedded into the tools people already use
• Value-creation opportunities exist now.
• Despite its flaws, gen AI can save time, reduce costs, and unlock new value. Holding off because the output isn’t perfect misunderstands the opportunity.
• Competitive advantage comes from using gen AI more strategically than others, not just faster.
• More perilously, your own customers and suppliers may disintermediate you by using it to take care of the tasks you previously performed for them.

• Where and When to Use Generative AI

   
• Rather than asking whether gen AI performs as well as a human, start by breaking down jobs into their component tasks and ask: Which of these is gen AI well suited to handle today?
• What distinguishes these examples from the first set isn’t necessarily the intelligence required but the cost of getting it wrong
• The suitability of gen AI for a given task depends not just on the capabilities of gen AI but on two deeper factors
• The first is the cost of errors:
• The second factor is the type of knowledge the task demands.
• Applying the Framework
• Rather than focus on the intelligence of gen AI (how smart it is and how fast it’s improving), organizations should examine its usefulness, which depends heavily on the task at hand.

• The no regrets zone.

   

  The lower-left quadrant, where the cost of errors is low and explicit knowledge is required, contains the clearest and most immediate opportunity for organizations. This is where gen AI should be deployed today and where AI agents will thrive in the future. Tasks in this quadrant rely on clear, documented data, and errors are relatively harmless. You don’t need perfect accuracy here. The real value lies in completing tasks faster, more cheaply, or at a greater scale than before.

• The creative catalyst zone.

   

  The upper-left quadrant, with a low cost of errors and a need for tacit knowledge, is where gen AI can serve as a creative catalyst, helping humans perform tasks that often benefit from originality. Crucially, the refinement of gen AI’s output and the final judgment on what to adopt rest with humans. Mistakes can be tolerated because the quality of the results is subjective: There is no definitive “best” marketing slogan or “perfect” product design because people’s views of what is best or perfect are personal.

• The human-first zone.

   

  The upper-right quadrant is where the stakes are highest. In this domain gen AI may act as an enabler but not a decision-maker. Tasks here involve subjective judgment, situational nuance, and complex decision-making—and mistakes carry serious consequences, whether financial, legal, reputational, or personal. Trust, ethics, and long-term strategy are often on the line. Errors can have lasting consequences:

• The quality control zone.

   

  The lower-right quadrant contains knowledge-heavy tasks that gen AI can technically perform well—because they are grounded in explicit, structured information—but for which even small mistakes could result in serious consequences. These are high-accountability domains such as law, finance, and software development, where information is clear and codified yet the standards for accuracy are extremely high.

• Anticipate the Impact on Your Industry

   

  The fact that your customers, suppliers, and competitors can access the same technology creates the paradox of access: Because everyone can use it, it becomes dramatically harder to capture value with it.
• Competitive pressure ultimately causes the gains to go to customers and suppliers through lower prices or better terms.

• AI-first entrants are coming.

   

  In the not-too-distant future your fiercest competition may not be your familiar peers but a new breed of solo entrepreneurs and micro-teams. Imagine starting a marketing agency today from the ground up. Rather than hiring dozens of people to conduct market research, write copy, design graphics, and answer questions from clients, a small team of experts (or even one intrepid entrepreneur) could eventually rely on AI for all these tasks.

• Customers and suppliers can use gen AI against you.

   

  Their access to gen AI can upend your bargaining power. Law firms have been dealing with a similar issue since the 1990s. Work that once required scores of paralegals and a complete law library could suddenly be done by one lawyer with an internet-connected PC.

• Building an AI-Based Competitive Advantage

   

  As we’ve noted, moving quickly is important, but speed alone won’t put you ahead of the impending competition.

• Mandate broad access to technology.

   

  Everyone in your company has tasks in all four quadrants of the framework, and so everyone has the potential to do more by using gen AI. Every single person in your organization should evaluate which tasks can be handled—better or even if just serviceably—by gen AI. Also have each person consider tasks that previously were too costly or time-intensive to do but that gen AI could perform inexpensively and quickly—for example, sending personalized holiday greetings to every business contact over the past year or summarizing every meeting attended.
• Start by removing the bottlenecks that keep these powerful tools out of the hands of your people. If access stalls at the IT desk or hides behind compliance forms, you cede ground to rivals whose staff can experiment in real time. IT departments understandably struggle to keep up with the relentless proliferation of ever-improving models and specialized applications.

• Ascertain where the data resides in your organization today and centralize it.

   

  All companies need to start centralizing data that has been scattered across or siloed in business units, functions, and geographies. Your infrastructure can anchor your competitive advantage.

• Identify the data that you aren’t yet collecting.

   

  Every activity of a business—from customer interactions to operational processes to internal emails and meetings—is a source of proprietary data to be tapped and leveraged.

• Redesign your organization.

   

  In the long term, it will not be enough to layer gen AI onto existing workflows. Organizations will need to redesign themselves around a gen AI–first vision of the business.
• Let’s look at data first. Even proprietary data eventually becomes commoditized. But it is hard for others to copy an organization that is tailored to continually exploit it.
• You also need to revisit how you get the most out of your people. Generative-AI tools free up chunks of time, but early research suggests that the windfall can evaporate into idle tinkering, low-value busywork, or outright downtime
• Managers should work with employees to estimate and track the hours AI shaves off their key tasks, set clear expectations for how those hours will be redeployed, and tie recognition or incentives to how effectively the saved time is used.
• In summary, strategic differentiation will come from three sources: (1) rapid and targeted deployment of gen AI across tasks, which is valuable in the near term if your competitors remain fixated on intelligence or paralyzed by concerns like hallucinations; (2) proprietary data that enhances gen AI’s performance or process fixes that prevent its value from being lost to organizational bottlenecks; and (3) unique people, processes, and culture—the “complementary assets” that make gen AI more valuable inside one organization than it is inside others.

32 more annotations...

• au fur et à mesure qu’une société évolue vers une plus grande complexité, les coûts de cette complexité imposés à chaque individu augmentent également, de sorte que la population dans son ensemble doit allouer une part croissante de son ‘budget’ (temps, attention, énergie, etc.) au maintien des institutions organisationnelles.

• chaque initiative collaborative introduit de nouveaux niveaux de complexité organisationnelle qui nécessitent des ressources dédiées pour être maintenus et coordonnés.

      
• L’organisation est plus complexe et nécessite un investissement considérable. Après des années à pousser cette évolution, le cabinet se rend aujourd’hui compte que la valeur ajoutée des nouveaux services ne couvre pas forcément les coûts de coordination induits, d’autant que ceux-ci ont été sous-estimés.
• elle crée un système de dépendances qui demande un entretien constant.
• Les réunions de coordination se multiplient, les canaux de communication se diversifient, et les processus de validation s’allongent.
• Cette ‘taxe’ collaborative ne représente pas juste un coût direct. Elle représente également un coût d’opportunité majeur : le temps consacré à la coordination est soustrait à la production de valeur directe.
• Plus une organisation développe ses capacités collaboratives, plus elle génère de procédures, de rôles spécialisés et de structures de gouvernance.
• L’entreprise se retrouve alors dans ce que Tainter décrit comme une spirale de complexité croissante, où elle doit investir davantage de ressources pour maintenir un niveau de collaboration qui devient de moins en moins efficient.
• La collaboration doit donc être abordée comme un investissement stratégique avec un calcul rigoureux du retour sur investissement.
• l convient d’identifier le point d’équilibre optimal où les gains de synergie compensent les coûts de coordination, avant que l’organisation n’atteigne son seuil de complexité critique
• Pour qu’elle devienne une réalité, la collaboration nécessite donc une approche disciplinée qui privilégie la simplicité des interfaces, la clarté des rôles et la mesure systématique de la valeur créée par chaque initiative collaborative.

9 more annotations...