The Emotional Labor in Exploratory Work
For all the creatives out there, and particularly early-career researchers.
What is exploratory work?
Here are some definitions:
- Work that comes with no instructions, blueprint, or template.
- What you do when you’re trying to produce something novel with a general goal but without concrete success criteria.
- Indefinite amounts of effort spent working on something that may be discarded in the final version of your deliverable.
However you define it, exploratory work is The Essential Work for any creative out there. For some, this could be composing music, starting a company, making a new recipe, designing an escape room, or writing a play. I am a computer scientist, so for me this is academic research.
Why is this necessary?
Since there is no blueprint, your path hasn’t been charted, and must be organically discovered. A musician doesn’t know what tune works until they’ve listened to many drafts that all sound bad for different reasons. Only by iteratively identifying what isn’t working can they make the right tweaks to get to their destination. If you’re familiar with gradient descent in machine learning, exploratory work is the equivalent of creating new data records in your dataset so you can calculate the error and backpropagate in the right direction.
Why am I talking about it?
Childhood, Computer Science, and Software Engineering
Let’s rewind to my childhood. I always loved learning new things as a child. I’d say curiosity was and still is one of my defining traits — anyone who’s met me can tell you that I love to ask tons of questions. Therefore, when I’d graduated from college and was pondering my career choices, learning and feeding my intellectual curiosity was a priority in any career path I considered.
At that point, I was a new grad software engineer at Meta building web and mobile apps. Having been programming for years before I started college, then having taken some 40 odd computer science classes, completed a slew of internships, work quickly became passé. Unlike in school, where every week in school I was learning something new, and getting to work on interesting math and programming puzzles, end-user application development hasn’t seen fundamental innovations in a long time — it’s efficient, streamlined, mundane, and suffice to say, wasn’t exactly feeding my soul.
In a matter of months, I began longing for challenging, intellectually stimulating work. It was at this point I felt called back to my undergraduate research days at Security and Privacy Research at Illinois (SPRAI), where I was in a community of people excited about continuously learning new things and sharing their knowledge. I was beginning to discover that once my brain realizes I’m starting to get good at something, it doesn’t want to keep doing it. Thus, I went back to research.
Challenges in Microarchitectural Side-Channel Attack Research
Over the years, computer architects have come up with a treasure trove of tricks to speed up processors. These tricks cause CPUs to exhibit distinguishable physical effects such as increased time or power consumption, as a function of the data being computed on. This can be used to leak secret information, such as cryptographic keys (check out GoFetch, a project I contributed to at Georgia Tech). However, since the code used to program CPU instructions (usually in Verilog / RTL) is proprietary, one needs to run a large number of benchmark programs to figure out exactly what trick gets activated on which pattern of instructions. It’s this opacity in processor design that makes it just like walking into the unknown, all alone.
Offensive security research can be an extremely competitive field, because the results have a practical impact instantly, and because software/hardware patches can render months of work wasted. Moreover, high-value targets (in the case of GoFetch, Apple M series CPUs) are being actively pursued by numerous groups simultaneously. This adds a lot of pressure to an already challenging activity.
However, pressure is not conducive to creative work: research shows you are at your most creative when you have plenty of unstructured time to brainstorm and let ideas flow freely, as opposed to when your nervous system is in fight-or-flight mode. Case in point, I was so afraid of failing, desperate to succeed, and unsure about how to proceed that it became paralyzing to make rational decisions and explore non-traditional solutions. Psychologically, it felt more tolerable to assume a given experiment wouldn’t work, than to try one with low odds of succeeding, similar to tunnel vision one has in stressful situations.
Why was this so hard for me?
While I was in an academic environment surrounded by people excited to discover the latest and greatest things in computer science, the grass is not always greener on the other side. Research, while requiring nearly constant thinking throughout the day, is rife with uncertainty. There are no instructions, so no one knows what to do to solve a problem, and so there is no one who can tell you what to do. This is actually the biggest shift I experienced, more generally in life, at the ripe young age of 22: the shift from being externally driven to (needing to be) internally driven.
I found this incredibly disorienting. Having spent 22 years mostly following other people’s instructions and becoming highly proficient at it, I had minimal experience understanding and embracing my own needs (particularly emotional; thank you Lindsay Gibson for writing my favorite book of 2024, Adult Children of Emotionally Immature Parents), and no idea how to spend my time when there was nothing concrete to achieve or required of me. In fact, I was highly risk-averse, and avoided anything that could fail.
This wasn’t just about not being able to enjoy exploratory games like Minecraft (I never understood what people got out of a game with no win conditions) but also holidays (which triggered massive uncertainty, made bearable by escaping into my favorite books), and of course, research. Life worked great when others needed things from me, but in the absence of hard, external constraints, such as deadlines, meetings, and appointments, I was quite lost. This brings us to emotional labor.
Emotional labor in research
Here are three things that make research emotionally demanding:
- You’re answering questions currently known by anyone and may be unanswerable.
- There is rarely a linear relationship between effort expended and progress.
- Often, you’re the only one in the world working on and stuck on a problem, which can be uniquely isolating.
These things combined mean that research can be extremely ambiguous, take as much of your time as you can give it, and yet provide little in the way of a satisfying feedback loop. This doesn’t even include the gauntlet that is publishing your research.
As a result, I found myself in a Catch-22; stuck between the discomfort of boredom (structured work, such as software engineering) and the discomfort of uncertainty (exploratory work, such as solving open problems).
The solution to all of this fell squarely in the realm of emotional health: regulating my (uncomfortable) emotions. I really wanted interesting work, but I also really wanted to succeed. However, as I’m sure is relatable to anyone doing research, most experiments fail, most of the time. Adapting a quote from Anna Karenina: all correct solutions are alike; each incorrect solution is incorrect in its own way.
Creative work doesn’t have short-term rewards — the reward is in the playing itself i.e. the process, and more specifically your experience and perception of it. If you’re stuck, you need to perform emotional labor to get comfortable with not knowing, which can be quite unsettling and demanding, especially since a lot of people doing research are already high-achievers used to one-shotting assignments.
Tangent: but I’m smart and knowledgeable!
You might wonder, what about intelligence? For the most part, smarter people can do more in lesser time. Surely, that should let you avoid these obstacles?
Not necessarily, because it is nuanced. In my experience, each type of intelligence requires intentional development, and cannot substitute for another.
Cognitive / analytical intelligence
This is something a lot of people spent years of their lives intentionally honing. It’s a significant part of why people in any sort of higher education program are admitted. Ironically, that also means that growing this has rapidly diminishing rewards. To advance the state of the art, you do need to be capable of getting to the edge of the field. However, once you do get to the state-of-the-art, I’ve found that only action has helped collect the data I needed to move forward, since the world isn’t set up to perfectly confirm your hypotheses.
From what I’ve observed, the more background knowledge and proficiency with methodological tools you have, the faster you’ll be able to get things done but also hit dead ends! You really can’t avoid dead ends altogether (Yoshi Kohno has a great article on this), since you don’t know what you don’t know. Interacting with this uncertainty reveals the missing information (often by trial and error) needed to move forward.
Emotional intelligence (literacy, attunement, and regulation)
This is another dimension of intelligence that is just as important and requires its own practice. If you’ve faced emotional neglect growing up, you probably have much less experience with it. Thankfully, it is possible to increase one’s emotional literacy later in life. In short, feelings are bodily sensations that last for minutes, serve as a vital source of information about our environment beyond what we’re cognitively processing, and help keep us alive!
So what?
Poor emotional regulation leads to addictions, anxiety, depression, and other unhealthy coping strategies (and downstream, a lot of undesirable things in this world). Good emotional health involves knowing your feelings (emotional literacy), noticing your own emotions (emotional attunement) and sitting with your feelings (emotional regulation), especially the uncomfortable ones. Easy to say, but hard to implement.
What do people do instead?
Avoid big challenging emotions, by:
- Denying them
- Repressing them
- Projecting their feelings onto others
- Intellectualizing as an escape
amongst a litany of other defense mechanisms. While in many cultures, women are encouraged to embrace and express their emotions, even in 2025, I can’t say the same is true for men, even though it is necessary for everyone.
Why?
Because emotions don’t go away. They remain in the body and manifest physically such as through psychosomatic conditions. In the words of traumatic stress researcher Bessel van der Kolk, The Body Keeps The Score.
Luckily, there are some great resources out there to build emotional fluency. I’d highly recommend these videos by Heidi Priebe, an amazing emotional health educator: Cultivating The Ability To Tolerate Emotional Discomfort, and Emotional Self-Intimacy: What It Is and How To Foster It.
Final thoughts
I think everyone can benefit from therapy, even preventatively. You may even think nothing is wrong and everything’s hunky dory, and I wouldn’t blame you for feeling that way. Sadly, mental ill-health is intangible and often invisible, making it tricky to track and treat.
In my experience, it is another form of worldwide inequality(!), stemming from cycles of generational trauma. Trauma creates an emotional blueprint and nervous system dysregulation that result in unhealthy (maladaptive) real-world reactions, and must be processed.
Trauma
A misunderstanding about trauma is that it is an incident, for example: experiencing war. It’s actually what happens within you psychologically in response to an incident. Here is a quick explainer.
Our value systems, perceptions of others and ourselves, ambitions, and mindset towards life, are rooted in those of our parents and childhood influences, entities that aren’t perfect. The way I see it, humans are literally programmed in their early childhood years using code that takes immense effort to modify later on, because of how strongly we grip to (what we think) makes us us.
As a result, many people have unprocessed childhood trauma (rewiring of behavior to adapt to toxic situations) that results in them internalizing limiting and distorted beliefs about themselves and the world, which don’t serve them in adulthood. These inhibit us from knowing ourselves and fulfilling our potential. Furthermore, these pervade all aspects of our lives, since they affect not only the lens through which we look at life, but also rewire our emotional and nervous systems. Thus, highly challenging and stressful situations such as research, can force them to come out in ugly ways, by straining inadequately equipped coping skills in emotional regulation.
I hope this article inspires you to learn more about yourself and improve your relationship with work. I invite you to extend yourself compassion as you embark on a deeply personal and discomforting journey. I’ve found that learning to perform this emotional labor has helped me find more peace and contentment. Research is not unlike life: you decide how to spend your time and what makes it worthwhile!
I originally titled this “The Emotional Labor in Speculative Work”, but as my CMU HCII colleagues kindly pointed out, “speculative” has many different meanings depending on what field you’re in, so I opted to use exploratory.
