The Hidden Architect: a Reflection on the work of Dr. Fei-Fei Li
The Quiet Labour of a Woman Shaping the World through Technology
It's the start of a new year, which means most of you just emerged from your hard-earned holiday break after surviving that special form of corporate torture we call annual planning (or quarterly, if you're a glutton for punishment like I am). I spent my November and December of 2024 in the same trenches - endless days of meetings, discussions, and debates mapping out our product's future. But something felt off this time. For days, I couldn't put my finger on what was nagging at me until it hit me: for the first time in five years, I was the only woman in the room. I found myself surprisingly unsettled by this realization despite being five months into my time at this new company.
That outlier feeling sparked a memory about one of my favourite podcasts — a season of You Must Remember This, a podcast about the forgotten history of Hollywood cinema. Specifically, I couldn’t stop thinking about the season that focused on Polly Platt, a product designer who was the hidden force that shaped New Hollywood cinema of the 1970s and 1980s. Platt was a woman knew exactly what it felt like to be the outlier in the room; but markedly deserved her place based on her talent and craftsmanship. Her fingerprints are all over masterpieces of the era, such as The Last Picture Show, Paper Moon, Pretty Baby and Terms of Endearment. She also mentored the next generation of auteurs like Cameron Crowe and Wes Anderson. She was labelled by Hollywood insiders as the “talent whisperer” - she’s the reason The Simpsons exists, having introduced James L. Brooks to Matt Groening. Personally I can’t imagine a world where I don’t have a quarter of my brain dedicated to being able to quote episodes verbatim. On paper, Platt was just a production designer. In reality, she was the hidden architect of an entire chapter of American cinema.
While sitting in those meetings, Platt's story echoed in my mind. I couldn't help thinking about all the women out there right now, sitting alone in boardrooms, their vision and ability to nurture raw talent going unnoticed. I found myself drawing strength from Platt's example. And as if the universe was trying to tell me something, I was simultaneously reading the memoir of another hidden architect - a person who’s labour has been critical to shaping the current technological revolution we’re experiencing. Dr. Fei-Fei Li, the godmother of AI, whose brilliance as a visionary, executor, and mentor mirrors Platt's in so many ways.
Li reminds me of Platt, though she's earned more recognition in her field: another quiet builder in a world that rewards the loud and boastful. She's a craftswoman (yes, I'm deliberately gendering that term) who's dedicated her life to a clear vision - what she calls her north star - of a human-centred future for AI. Something about these two women's stories wouldn't let me go. What makes someone a hidden architect? It's more than technical excellence - it's about having the audacity to pursue a vision others can't yet see, and the patience to build it brick by brick, far from the spotlight.
What Makes a Hidden Architect
I discovered Dr. Li by listening to her on Kara Swisher's On podcast in late 2023, right before her memoir was release. If you know anything about tech journalism, you know Swisher doesn't suffer fools gladly. She's got this razor-sharp ability to slice through the tech industry's smoke and mirrors, separating the real builders from the hype merchants. After that interview she continued to sing Dr. Li’s praises regularly. That tells you everything. If I’m right and most of you are unfamiliar with Dr. Li, let’s summarize her career in a snapshot: she is someone who pioneered major breakthroughs in the field of computer vision and machine learning through her work on ImageNet, a vast visual database that became foundational to modern AI development. Not only that, she quietly pushed for making AI more inclusive, more ethical, and more humane through her work at Stanford's AI Lab.
Diving into Dr. Li's memoir, The Worlds I See, I had one of those spine-tingling moments of recognition. In her story, I saw Platt's ghost - the same three qualities that define these hidden architects: that almost desperate need to pursue a vision nobody else quite gets, the stomach-churning determination to keep pushing when everything and everyone says stop, and this profound commitment to pulling others up alongside you. Let's unpack how these traits showed up in Dr. Li's journey.
Vision
If you work on AI products, you know the critical importance of training data quality. But this wasn't always a given twenty years ago. Dr. Li's fundamental insight - that quality data was the missing half of the AI puzzle - challenged years of academic orthodoxy that the algorithm was most important. For decades, it was newer models and math that commanded the lion's share of research funding and academic attention. Yet back then, research continued but the results continued to be underwhelming.
My own entry into tech intersected with this evolution, though in the realm of language rather than vision. At my first startup, I worked alongside PhDs developing recommendation algorithms for news and media websites. I spent countless hours in our small company server room (yup, before the cloud) manually training and classifying data, an experience that makes Dr. Li's vision all the more resonant to me personally.
Dr. Li's conviction about the primacy of data in computer vision came with a sobering understanding: it would require an immense investment of human labor to classify images at scale. The initial estimate for building ImageNet to the necessary scale was nineteen years - even with a team of poor graduate students at the ready to take on some side work. This timeline, daunting as it was, crystallized a truth about technological breakthroughs: they often demand a level of sustained effort that seems almost irrational at the outset.
Grit
Vision alone, however compelling, doesn't guarantee transformation. What distinguishes Dr. Li's narrative is the relentless persistence required to manifest her vision in a field that wasn't ready for it. As funding dwindled and peer skepticism mounted, a pivotal insight emerged through her lab's collective problem-solving: Amazon's Mechanical Turk could fundamentally reshape their approach to image classification. This platform, offering human validation for mere cents per image, accelerated what had seemed an insurmountable timeline.
But when it was completed, what next? Academics in the AI field were not looking for this kind of a dataset; not yet anyway. Dr. Li’s lab had accomplished this big feat of classifying and annotating over 14 million images and no takers. So they launched a competition, ILSRV (ImageNet Large Scale Visual Recognition Challenge). ILSRV transformed perceived indifference into engaged competition. What began as a tactical move to attract researchers evolved into the definitive benchmark for computer vision progress. It fostered innovation and accelerated progress in AI research — including the seminal year when AlexNet won the competition in 2012.
AlexNet, developed at the University of Toronto by Alex Krizhevsky, Geoffrey Hinton, and Ilya Sutskever, represented more than technical achievement. It vindicated two parallel streams of conviction: Hinton's unwavering belief in neural networks despite their computational demands, and Dr. Li's insistence on the primacy of quality data. The convergence of affordable GPU processing power and ImageNet's meticulously curated dataset produced what Dr. Li describes with elegant precision:
“Data, it seemed, had a way of breathing fire in the system…All along, neural networks hadn’t needed fancier math and more exotic abstractions. They were simply waiting for a clearer picture of the world we expected them to understand.” - The Worlds I See, 226.
Breathing fire indeed. The AlexNet team's trajectories illuminate the breadth of AI's evolution: Hinton's theoretical foundations earned him the 2024 Nobel Prize in Physics, Sutskever helped birth OpenAI in 2015, and Krizhevsky, after selling his company to Google, continued his research -he out of the three embody the archetype of the hidden architect the most.
Mentorship
The reverberations of ImageNet extend far beyond technical milestones. Dr. Li's impact manifests most profoundly in the careers she's shaped and the minds she's influenced. Through SAIL, she cultivated a generation of AI innovators who are now reshaping our technological landscape. Consider Andrej Karpathy, whose work on image captioning laid foundations now ubiquitous in content creation tools. His trajectory - from SAIL to OpenAI, Tesla, and back to OpenAI - exemplifies how computer vision expertise transforms seemingly disparate domains, from autonomous vehicles to creative AI applications. Equally significant is Timnit Gebru's journey. Their collaborative research using Google Street View data to analyze neighbourhood demographics demonstrated AI's potential for sociological insight. That Gebru later emerged as a crucial voice in AI ethics, challenging Google's approach to the rapid development of their own LLM, speaks to the complex interplay between technical innovation and ethical responsibility - a tension increasingly central to AI development.
Dr. Li's commitment to expanding AI's horizons culminated in AI4ALL, co-founded with Olga Rossavosky. It’s a two-week AI crash course to help introduce young curious minds into the world of AI, specifically focusing on bringing in new diverse voices to the field. In a field where data inherits societal biases, diversifying the next generation of AI architects becomes not just morally imperative but technically essential for building more comprehensive and equitable systems.
Reading Dr. Li's memoir left me with that rare feeling of recognition; when someone articulates something you've felt but couldn't name. I had read the blueprint of what makes a hidden architect: unrelenting vision, quiet determination, and perhaps most importantly, the grace to reach back and pull others forward. But her story isn't an isolated case.
Dedicated to the Other Hidden Architects
There are more hidden architects out there than we realize, quietly reshaping our digital world while others grab headlines. I had planned to write about two more women who exemplify this - their stories deserve full newsletters of their own. But let me give you a glimpse, because they perfectly capture both the power and the challenge of being a hidden architect in today's attention-hungry tech world.
My husband's daily digital habits led me to these stories. Two products he can't live without - YouTube and Facebook Marketplace - were shaped by hidden architects who knew when to work in the shadows and when to step into the light. The late Susan Wojcicki saw YouTube's potential when others thought she was crazy for the $1.65B acquisition price (in 2024 it surpassed $50B in revenue). She turned chaotic video platform into a sustainable business; not only for Google but for the creators that used it. I doubt that the thriving creator economy we have today would exist without her quiet labour. Then there's Deborah Liu, the mind behind Facebook Marketplace. As a new mom, she turned the idea of second-hand market of mom groups into something that Facebook could productize. It took years of pitching, building, iterating, failing, and navigating immense challenges (read the fully story in Lenny’s Newsletter).
Here's what I've learned from these women: there's immense power in being underestimated, in being able to focus on the work above all else. But I've also learned that staying completely hidden can mean your vision never reaches the scale it deserves. The most effective hidden architects I've observed have found that delicate balance - knowing when to stay heads-down in the craft and when to step forward and tell their story, not for ego but for impact.
I know many more hidden architects in my own career - women who've mastered this dance between visibility and focus. If you think of someone worth highlighting, someone who's found that balance between building and being seen, drop in a comment or email me (stacey[at]nofearproductleadership.com). I'm thinking about making this a regular series, not just to celebrate these builders, but to help all of us navigate the tension between craft and visibility in our own careers. Sometimes the loudest voice in the room isn't the one making history - it's the quiet builder in the corner, head down, hands dirty, knowing exactly when to stay hidden and when to be seen.