Make No Little Plans: Why You Should Think Big in the AI Era

A Quote That Sparks Reflection

Recently, a brief tweet resonated widely across the tech community — "Make no little plans." This phrase originally comes from the renowned American architect Daniel Burnham, whose complete statement reads:

"Make no little plans; they have no magic to stir men's blood and probably themselves will not be realized."

It's worth noting that Burnham (1846–1912) was one of the most influential urban planners in America at the turn of the 20th century. He led the master plan for the 1893 World's Columbian Exposition in Chicago and authored the famous Plan of Chicago (1909) — the first comprehensive urban master plan in American history, which profoundly shaped the development of modern urban planning as a discipline. This quote was born from that grand practical context, reflecting Burnham's deep understanding of the relationship between scale and vision: only goals ambitious enough can mobilize sufficient resources, talent, and social will to ultimately drive real transformation.

In today's rapidly evolving AI landscape, this statement — born over a century ago — is more relevant than ever.

Why "Make No Little Plans" Matters Especially in the AI Era

The Technology Inflection Point Has Dramatically Lowered the Barrier to Grand Plans

In the past, "grand plans" typically meant massive funding, large teams, and lengthy development cycles. But AI is fundamentally changing this equation. The maturation of large language models, generative AI, and automation tools now enables individuals and small teams to tackle project scales that were once imaginable only by large corporations.

The current "technology inflection point" in AI is driven by several key breakthroughs: the Transformer architecture proposed by Google in 2017 laid the foundation for large language models; OpenAI's GPT-3 released in 2020 first demonstrated the Emergent Abilities of large-scale pre-trained models; and the launch of ChatGPT in 2022 brought these capabilities into the public spotlight. Meanwhile, the thriving open-source ecosystem (such as Meta's LLaMA series, Mistral, etc.) has significantly lowered the barrier to accessing models, and the widespread availability of cloud computing infrastructure means compute power is no longer a wall for small and medium teams. The maturation of AI coding assistants (GitHub Copilot, Cursor), generative design tools (Midjourney, Stable Diffusion), and automated workflow platforms together form the practical foundation of "technology leverage."

A single person using an AI coding assistant can complete in weeks what previously took months of development; a startup team of three to five people can build product prototypes rivaling those of tech giants using open-source models. The amplification effect of technology leverage has caused the cost of "thinking big" to drop dramatically.

The Hidden Risk of Small Plans: Why Small Goals Are Actually Harder to Achieve

Burnham's insight was this: small plans not only lack the power to inspire — they are also inherently difficult to realize. This sounds counterintuitive — aren't small goals easier to achieve?

Modern psychology and organizational behavior provide strong support for this paradox. Stanford psychologist Edwin Locke's "Goal-Setting Theory" demonstrates that high-difficulty, high-specificity goals tend to produce higher performance than low-difficulty goals, because they activate stronger cognitive engagement and sustained motivation. Google's widely adopted OKR methodology also has a famous principle: if your goal completion rate is always 100%, your goals are set too conservatively. At the talent attraction level, research shows that top engineers and creative talent prefer to join projects with a "sense of mission" rather than teams pursuing incremental improvements.

In reality, small plans often fail to attract sufficient talent, resources, and sustained motivation due to their lack of appeal. In the AI field, this is particularly evident: projects aiming for "minor improvements" are quickly drowned out in fierce competition, while teams that dare to propose disruptive visions — whether OpenAI's goal of artificial general intelligence or Tesla's full self-driving vision — even if not yet fully realized, are pushing the entire industry forward.

The "Gravitational Effect" of Grand Visions

A sufficiently bold plan creates its own gravitational field: it attracts like-minded talent, draws investor attention, and builds user anticipation. This positive feedback loop is something small plans can never create.

OpenAI was founded in 2015 with the founding mission of "ensuring that artificial general intelligence (AGI) benefits all of humanity" — a goal that many industry insiders at the time considered a far-fetched sci-fi fantasy. It was precisely this grand mission that attracted top researchers and secured early supporters' commitments. Tesla's Full Self-Driving (FSD) vision works the same way: despite repeatedly delayed timelines, this goal continues to attract the world's top computer vision and reinforcement learning talent, driving a technology race across the entire autonomous driving industry. Both cases confirm a pattern: grand visions, even when not fully realized, can propel industry leaps forward, and their intermediate results alone carry enormous value.

Looking back at major breakthroughs in AI, nearly every one originated from some "unreasonable" grand vision — making machines understand human language, having computers generate photorealistic images, making AI an intelligent assistant for everyone. These goals all seemed overly bold when first proposed, but it was precisely that boldness that drove genuine innovation.

How to "Think Big" with AI in Practice

Start with End-State Thinking, Then Work Backwards

Don't ask "what can current technology do?" — ask "what should the ideal state look like?" and then work backwards to map out each step of the path. This methodology has two important corresponding concepts in Silicon Valley: one is Elon Musk's celebrated "First Principles Thinking," which means discarding existing constraints and conventions to re-derive solutions from the most fundamental facts; the other is the "Moonshot Thinking" advocated by Google X laboratory.