This month we’re discussing Scott Hartley’s The Fuzzy and the Techie in the Tech Leaders Salon. Below are some preliminary ideas. I will be publishing a more detailed reflection after the discussion here.

First off, a few words about why I chose this book. I first read it in the summer of 2022. Back then, I was doing research for a course about the importance of asking questions in identifying the right problems to solve. In a way, this book was the starting point that helped me crystallize the idea and catalyze the process for writing my book.

While Hartley’s book came out in 2016, the points he makes are more relevant than ever, especially with the surge in AI technology. Given that in August we will be discussing The Alignment Problem, which examines some of the challenges that present themselves to us when building AI systems, and how we can do so in a way that’s more aligned with our values, I thought that The Fuzzy and the Techie could be a good entry point and base for the discussion.

The premise is quite simple, and there’s hardly anything radically new there. When all else is equal, like technical skills, capital, and good teams, the differentiating factor in today’s tech-driven economy is how we approach and solve problems, improve our decision-making, cultivate creativity, build better products and services, communicate and interact with each other, and design a human-centered world. This requires developing a robust background in the liberal arts and social sciences, and can be achieved by bridging the gap between technical and soft skills.

While the point may not be all too surprising for you if you find some value in the liberal arts, the book does offer a compelling exploration of how the marriage between those with a science background (STEM, or techies) and those with a liberal arts foundation (fuzzies) would lead to building technology that solves pressing problems, augments and improves our lives rather than just building technology for its own sake.

In what follows, I’ll share a brief overview and notes on the book, including remarks and quotes.

The Chasm

I always joke that philosophy is useless. The long story short of this is that if measured by STEM criteria, it indeed is useless; the results of reading philosophy are not immediately visible in the same way as learning how to code is. Despite its uselessness in that sense, I do still think philosophy, and the liberal arts in general, are quite useful.

Hartley also brings this up at the beginning of the book. At Stanford University, the labels techie and fuzzy were used to describe STEM students and liberal arts students respectively. However, “beneath these lighthearted appellations rest some charged opinions on the relative value of each type of degree, on the importance of the direct vocational application of a college degree, and on the appropriate role of education.”

In other words, liberal arts education is useless, and only STEM education is valuable. This view spilled over into Silicon Valley, where the emphasis is usually on practical fields of study, pragmatic results, and coding above all else.

This chasm is based on a false dichotomy and opposition, according to Hartley, because both worlds need and depend on each other.

What happens, though, when the tech stack and tools become more accessible and democratized than ever? In response to this question, Hartley writes:

“As technology offers an ever more accessible toolbox, our differentiation—our very competitive advantage—becomes the very thing liberal arts programs teach”.

The book contains innumerable interesting examples that support the points Hartley makes in each chapter. I will only list one or two that serve as a good illustration.

I’d like to focus on two major themes: 1) The human factor in technical and data-heavy contexts. 2) Reimagining the future of education and jobs in an AI world.

Full Stack Integrators and Asking Better Questions

In a democratized tech world, we stand before a new reality. Being technical, having a deep grasp of engineering or coding used to provide people with a competitive edge. What happens when the technical barrier is no longer there? Pieter Levels shared a Tweet a few days ago about this; he writes: “I increasingly have no idea about anything anymore, time is changing too fast, and with AI everyone is equal now, nobody really knows anything while there is a superintelligence that’s actually moving things into a direction that nobody can predict.”

My answer to this emerging crisis is to read more philosophy. Hartley also agrees, but offers a more well-rounded, non-prescriptive solution to this, albeit one that takes some effort! Now you can easily, and in record time, bring your idea to life with AI, whether through vibe coding or a quick prototype.

Once the idea materializes, you still need to market it, raise funding if needed, convince potential customers to buy into it, and then decide how to grow the product or service. Even before doing any of that, the speed and rate at which we can ideate and develop a product make it more essential for us to ponder the reason why we’re building it, and what problem we’re solving.

This entails a shift from being full stack developers to full stack integrators, combining skills that go beyond the mere technical know-how to include communication, problem-solving, decision-making, and understanding human nature and psychology.

“‘Full stack developers’ have given way to ‘full stack integrators’”.

This shift is illustrated by the example of Katelyn Gleason, a theater arts major who founded the health tech company Eligible, which automates patient insurance claims, making it easier for hospitals and doctors to verify them. Gleason acquired her industry knowledge after working briefly in a health care startup. She identified an inefficiency in how insurance claims were done, taught herself how to code, had other freelance engineers help her build the prototype, and managed to raise $25 million in venture capital. She explains that she managed to pull it off because of the skills she acquired at theater school:

“In theater, the playwright gives you the play, but you have to tell the story. . . I knew I just had to figure out how to tell the right story. When you start rehearsal, you’re completely lost. You don’t know the characters at all. When you start to build a product, when you start to build a company, and you don’t even know what your product is going to be, it’s the same feeling. You’re completely lost. I learned in the rehearsal process that if I worked hard enough, I could gain that internal clarity where I would start to take off like a rocket ship.”

In another chapter, Hartley also challenges the narrative that big data alone makes human theory or intuition obsolete, emphasizing that “data do not speak for themselves; we need smart questioners”.

Big data in and of itself is not enough; it doesn’t always tell the full story, and calls for contextual or industry knowledge, interpretation, and human insight. Also, sometimes the algorithms we build are susceptible to our biases. Since these issues will also be brought up in the August Salon, I’ll just focus on one interesting example.

Hartley argues that the smart questioners provide a social, ethical, and geopolitical context that the data alone lack. This approach was followed by the Good Judgment Project led by Philip Tetlock. Competing in a US intelligence challenge, they used a diverse team ranging from Navy officers to history majors to apply human context to raw data. While the algorithms and tracking system trace ships as mere dots, the analysts add a historical and geopolitical context to the data, asking questions about the reason why these ships are moving in a certain way and are present in a certain location, avoiding false alarms, unnecessary standoffs, and rising political tensions.

“Computers don’t detect novel patterns and new behaviors; humans do: Humans, using technology, testing hypotheses, and searching for insight by asking machines to do things for them.” “The imperative is not to figure out how to compute . . . but what to compute. How do you impose human intuition on data at this scale? We start by designing the human into the process.”

Human Judgement, The Future of Work and the Fuzzy Mindset

Interestingly, some of the problems and questions tackled in the book keep recurring. I recently read an article about AI, wine apps, and wine experts. The article cites different studies that were conducted to determine what the best wine rating system is, and whether AI could do it better.

The takeaways were that while AI systems excel at identifying wines based on excellent chemical analysis and pattern recognition, they cannot replicate the embodied experience of smell, taste, and touch, which are necessary for a more informed and consistent recommendation based on people’s preferences. The article argues that AI ought to serve as an augmentation for wine experts who provide better judgement and explanatory abilities, rather than completely replacing them with AI.

Hartley provides a similar argument. He distinguishes between artificial intelligence (AI), which in fact seeks to replace human ability, and intelligence augmentation, which focuses on a supplementary and symbiotic partnership with the machines. This is what clothes retail company Stitch Fix, founded by economics major Katrina Lake, does. It uses machine algorithms to filter inventory, but ultimately relies on human stylists to make the final selection.

With more and more automation, algorithms, and AI taking over some of the tasks we do, the fear of becoming obsolete is justified. We don’t really know what the jobs panorama will look like 1, 2, and 5 years from now.

The way Hartley looks at it is from a routine vs non-routine perspective. He notes that while machines do in fact excel at routine tasks, and we can say that AI now excels at some more complex tasks, they eventually struggle with non-routine complexity. These include interpersonal skills like coordination, negotiation, empathy, and social perceptiveness, or simply the ability to read the room. Such skills are and will become more of a vital economic force.

“Social skills act as a kind of social antigravity, reducing the cost of task trade and allowing workers to specialize and coproduce more efficiently”.

To that end, Hartley suggests using the Cynefin framework (which I’m pretty sure Laksh can do a much better job explaining) developed by Dave Snowden, which sorts situations into categories like simple (known knowns), complicated (known unknowns), complex (unknown unknowns), chaotic, and disorder. The determinant of each situation is the underlying cause-effect relationship.

In cases where this is easy to discern (simple), decision-making is straightforward and can be codified.

When the cause-effect relationship can be discerned but calls for expert analysis (known unknowns), decision-making needs further sense-making. In both of the previous cases, right answers exist.

When the cause-effect can be determined only in retrospect (unknown unknowns), a right answer is difficult to determine beforehand. In this case, the best approach would be to let the patterns emerge, and think about the decision-making process in terms of best bets.

In situations that are chaotic and disorderly, the best approach is to either stop the bleeding and take actions that would reestablish order, or to break the situation into individual parts and try to approach each based on the same framework above.

Either way, different contexts and situations demand a host of skills that need human judgement and intervention, specifically when the situations are complicated, complex, chaotic, or disorderly.

“Dynamic, nonroutine, highly agile interpretation and subsequent improvisation of analysis and action”.

How can we prepare ourselves for such a new reality where AI takes over the routine tasks, leaving us with non-routine ones? Hartley argues that “the fuzzy combined with the techie is the formula for the most transformative, and most successful, innovations”.

He advocates for a philosophy where our education should always be in beta, adopting a curious mindset in pursuit of a blend of skills including soft and technical ones. This is a lifelong process that also means rethinking our educational systems and processes.

In addition to more holistic and blended liberal arts and technical programs, Hartley also highlights the importance of an active learning model, quite similar to the Socratic method, where teachers act more as guides than strict lecturers. As such, cultivating a curious mindset and building a mix of soft and technical skills make it easier for us to approach education as a lifelong process of learning and learning how to learn.

“It’s not about concrete knowledge, it’s about higher order thinking skills, and the ability to perpetually learn and grow.” — Diane Tavenner

And on that note, I’ll close this overview with two more liberal arts quotes:

“Technology alone is not enough—it’s technology married with liberal arts, married with the humanities, that yields us the result that make our heart sing”. — Steve Jobs

“The liberal arts are still relevant because they prepare students to be flexible and adaptable to changing circumstances.” — Georgia Nugent

However, as I mentioned recently in a tweet: the ‘we need the humanities more than ever to thrive in the age of AI’ is becoming the new ‘we care about the environment’ corporate social responsibility stuff. No need to do anything about it, just talk about how truly important it is, and score more ‘oh you’re cool’ points.