Not Comprehending Subtraction To Improve Efficiency

Article published in the January 2025 Newsletter of the Global Smart Grid Federation.

Some of my first learnings on human biases in comprehension began with the challenges in formulating energy efficiency programs. After 25 years, it is still true today. From design to roll-out, comprehending “less is efficient”, does not ring well with most people (consumers and policy makers alike). It gets noticed only when a big news event shakes everybody up.?

?The recent news about China’s launch of DeepSeek R1 Open AI models at a fraction of the cost of ChatGPT, is a classic display of how efficiency pays. DeepSeek’s efficiency gain rings true in all aspects (a) less computing power (hence energy); (b) less powerful chipsets (optimal use); (c) small sacrifice in accuracy from 32 to 8 decimals (75% less memory); and (d) having only 37 billion parameters activated out of 371 billion (rest available on call). From top to bottom and inside out, this effort points to painstaking improvement in efficiency.

?Yet, the concept of embedding efficiency in systems design is not seen by many as being important. Falling costs often lead to inefficient use. Most companies continue to innovate using “more is power” principles, as evidenced by hyperscale data centers, giga factories and huge warehouses, all requiring large power, cooling and backups. This also extends to transportation, housing and personal gadgets. Even green power (solar, wind, storage, etc.) are sacrificing efficiency with falling costs. With each innovation and upgrade, efficiency takes a back seat, in the singular pursuit of speed and power.

?When I used to teach two continuing education courses, Management of Innovation and Finance for Engineers, at the University of Toronto years ago, I would often cite examples of corporate focus on top line metrics such as income, revenue and asset growth instead of efficiency metrics such as rates of return, cost of capital, and debt service ratios. Even today, financial market reporting focuses on “additions” (revenues, assets, market share, staffing) while out shadowing efficiency metrics such as ROA, ROIC, and employee efficiency. In short, corporate “growth” is perceived to be more progressive than corporate “wellness”.

?The above (“more is power”) pursuit would not be possible if consumers made choices focused on cost and efficiency in their product selection and product life cycle (low energy footprint, recyclability, and supply chain circularity). Perhaps, we humans are “pre-wired” to not easily comprehend “subtraction” (reduce, shrink, lower, condense, shed) as being beneficial. We are more apt towards “accumulating/building/buying”, but poor at “disposing/donating /divesting”. This is true in our investments, professional progression and business objectives. We are good at “creating” (addition) but not so good at “rationalizing” (subtraction).

?This human bias is evidenced in all aspects of our lives, such as:

?1.????? Health – inability to combat marginal weight loss, blood-sugar, cholesterol, through diet control (“subtractions”) but combating these by taking medications (“addition”).

?2.????? Economics – emphasizing progress through the lens of (a) gross GDP (over per-capita); (b) output tonnage (over efficient production); and (c) gross employment (over human capital efficiency). The economic grandeur lies in addition.

?3.????? Energy – planning using (a) capacity additions (over demand reduction); and (b) gross energy consumption (over energy intensity). Reducing energy footprint is seen as becoming poor.

?4.????? Urbanization – emphasizing the need for growth driven by (a) economic output (over agrarian efficiencies and food security); and (b) infrastructure capital growth (over unsustainable per capita maintenance costs). Cities with more amenities are better than small agrarian towns and villages.

?This efficiency conundrum (involving subtractions) is a big hurdle to sustainable global progress. From the rich to poor nations, this message is simply not getting through to its people. It is ironic that even those countries who can ill-afford, are not focusing on efficiency in formulating their policies, standards and codes. Consuming less is somehow considered a sacrifice, inconvenient, or being poor.

?The recent COVID crisis saw versions of this stalemate as nations debated over equitable input allocation and need-based distribution. The current climate change dialog is also slipping back on account of this. As a global society, we need to solve this naivety in “subtraction” in all aspects of our public, institutional and private lives.

?Some would argue that levies and taxes are all “additions” imposed to reduce consumption. Yes, in principle, but are not having any substantive result in efficiency improvements. We need to explore why this is the case.

?Perhaps, we are indeed pre-wired not to comprehend subtractions. Will the DeepSeek R1 team teach us this!