The changes brought about by artificial intelligence, both in speed and scale, far exceed those of any previous technology. People say things like, “The competitive moats that companies spent years building can become fragile with sufficient computing power,” or “Most white-collar jobs will be replaced by AI.”

As the media amplifies such narratives, fear has become the dominant emotion. Companies are rushing to adopt AI, afraid of falling behind. Employees, meanwhile, are embracing all kinds of AI tools, often setting aside concerns about security and privacy in search of a sense of career security.

Like a shot of adrenaline, fear can prompt hasty action, but it cannot sustain a long journey. A more enduring source of motivation is a clear sense of purpose and value that enables companies and individuals to go further. How can companies break out of this classic prisoner’s dilemma?

When seeking to implement AI, many companies still rely on the same approach they used during the previous digital transformation wave: benchmarking against competitors and identifying the most efficient pain points to address.

This outward-looking methodology can be effective in a relatively stable business environment. Companies can observe what leading players are doing, then adapt and optimize those practices for their own circumstances. Take e-commerce as an example. Around 2009 and 2010, brands such as UNIQLO and Philips opened flagship stores on Tmall. As the e-commerce model matured, more and more companies learned from these early movers and gradually built their own e-commerce capabilities.

But in the age of AI, looking to others for answers can feel like seeing through a haze. Companies may not know how close those examples really are to their own situation, nor whether what they are seeing is even real.

What companies truly need is a stronger internal compass.

Start with value, not AI

The current enthusiasm around GEO (Generative Engine Optimization) offers a good example.

At first glance, the logic behind GEO seems flawless: as consumers increasingly turn to AI assistants for information, brands should seek to gain an advantage in AI-generated recommendations, just as they once competed for search rankings or exposures within recommendation feeds.

However, there is a critical flaw of the premise – a series of short-term actions can influence how a model perceives a brand given AGI is within reach.

What large language models ultimately do is distill the collective understanding of a brand into an answer for the user. The core task for brands, therefore, is not to optimize for the model first, but to define their brand essence from within, by articulating their value in the language consumers actually use, and continuously building that perception through effective channels. Only then does it make sense to ensure that large language models can accurately understand and express it.

This is where companies need to look within.

The core value of a company or brand is, in essence, the embodiment of human motivation within a business organization:

  • Why does the company exist?
  • What does it want to create for its customers?
  • What distinctive perception does it want to establish in the market?

Only when these answers are clear can a company expect AI to understand and represent it accurately.

The same principle applies to the ongoing hype around AI-driven efficiency.

While AI dramatically lowers the cost of gathering and processing unstructured data, and accelerates content creation across the board, the starting point for business evaluation should not be “What can AI do?”. Companies must instead return to their core value chain and apply first-principles thinking to ask the foundational questions.

On information and decision-making, the first question is not “How much insight can AI generate?” but rather “What information actually drives better business decisions?”. If the problem itself has not been clearly defined, more information simply means more noise.

On operational efficiency, instead of hastily asking “Which tasks can we outsource to AI Agents?”, companies should first map out their workflows. High-frequency, repetitive, and rule-based tasks rarely require sophisticated AI: traditional automation gets the job done. If a business hasn’t automated these basic tasks yet, it’s too soon to jump to AI agents.

Where agents truly shine is in scenarios that demand contextual judgment at scale, such as tailoring client sales proposals, curating localized training for retail teams, or evaluating potential store locations.

Regarding knowledge and communication, before deploying generative AI to produce reports, sales decks, or marketing collateral, leadership must diagnose the root cause of friction. Is the information difficult to source, disorganized across teams, or simply presented ineffectively? Internal bottlenecks usually stem from the first two issues, while external communication gaps tend to stem from the latter. Ultimately, structuring enterprise knowledge and upskilling employees in AI collaboration offer a far higher ROI than merely stacking agentic tools.

Root yourself within, then reach outward

For companies navigating change, understanding one’s core value is the most crucial benchmark, serving as the foundation for all future transformations.

By applying first-principles thinking to the entire value chain, companies can leverage technological innovation and expand their ability to create long-term value.