How to find Creative Talent?
Open questions on book Talent by Tyler Cowen & Daniel Gross
Hi, I’m Aastha, and on this newsletter I write essays on books, education, human creativity, and more. I have another newsletter Live Longer World, where I interview scientists researching the frontiers of longevity science and write about health & longevity practices.
The following essay explores unanswered questions and thoughts I had upon reading the book Talent: How to Identify Energizers, Creatives, and Winners Around the World by Tyler Cowen and Daniel Gross. While the essay might read as a critique, I write it merely as an open discussion of ideas for the sake of intellectual rigor and curiosity.
Brief Overview of the book
The main question the book seeks to answer is: “How do you find talent with a creative spark?”
The crux of the book is about providing readers with tangible methods for identifying “creative talent”. The authors don’t write about any kind of talent, they specifically write about “creative talent”, and this is important to keep in mind upfront.
They believe that traditional bureaucratic approaches for identifying creative talent are “deadly”, and “are poorly serving the American economy” as it “prizes consensus above all else” and “stamps out individualism.” Hence, we need better ways of identifying such talent and the book seeks to provide some.
Taking a step back – what is “talent with a creative spark” or short for “creative talent”?
What is “talent with a creative spark”? The authors say: “In referring to the creative spark, we mean people who generate new ideas, start new institutions, develop new methods for executing on known products, lead intellectual or charitable movements, or inspire others by their very presence, leadership, and charisma, regardless of the context.”
It’s clear that the authors are not addressing those hiring for non-creative roles. What is not so clear is whether the authors are discussing strategies for identifying creative talent regardless of the specific role vs. identifying creative talent for a specific role.
Identifying talent regardless of the specific role is what I call open-ended creative talent search. Conversely, identifying talent for a specific role is what I call closed-ended creative talent search.
While I think the authors primarily discuss strategies for open-ended creative talent search, they do not make this clear1. As a result, someone reading the book and looking to adopt their strategies to closed ended talent search like hiring a creative person for a specific role (i.e. a product designer) could be misled. This is because most people hiring for creative roles have to take into account role-specific creativity or what I call closed-ended creativity.
Before we dive into the book’s strategies for talent identification and where they fail, let’s take a closer look at the two modes of creative talent search since this distinction is key to the talent identification decisions that should be made.
Open-ended vs. Closed-ended Creative talent search
Filling roles for most creative jobs is a closed-ended creative talent search. You want the most creatively talented person for the specific job at hand. You're not hiring the person to be successful “regardless of context” over their lifespan at the company. You're typically hiring the person to be successful at their specific role. For example, if you identified a creatively talented writer, you wouldn't want them in the role of a designer.
Conversely, open-ended creative talent search is making a bet on the person, regardless of the specific project. You search for creatively talented people and make a bet that they will be successful over the arc of their career. Early-stage startup investing often takes this form as early-stage venture capitalists bet on the founder, not necessarily the idea.
The authors don’t make this distinction between open and closed-ended creativity very clear. Hence, someone applying their tactics to closed-ended creative talent search could be left with a lot of questions.
As part of their work, both Daniel and Tyler are engaged in open-ended creative talent search. Daniel is an early-stage startup investor through his company, Pioneer, and Tyler awards grants to talented individuals through Emergent Ventures. I'm sure they consider the project or the idea the person is working on, but I believe the emphasis is on supporting talented individuals. This might be why they missed the importance of distinguishing between open and closed-ended creative search.
Additionally, both Tyler and Daniel are looking for big wins, and don’t care if there are small mistakes. However, in most closed-ended creative talent search, the mistakes do matter. It’s not as easy to fire a bad talent hire and the costly mistakes can add up.
Having established this distinction between open-ended and closed-ended creative talent search, let's look at how the search and identification process for the two types differ.
Talent identification: Search vs. Search Optimization problem
The authors imply that talent identification is a search and identification problem. However, this is more nuanced. Open-ended creative talent identification is a search problem. But closed-ended creative talent identification is a search-optimization problem.
Unlike open-ended creative talent search which focuses on identifying creatively talented individuals, closed-ended talent search is a two-pronged problem: (1) Find a creatively talented individual (2) Make sure the creative talent aligns with the job at hand. What makes this problem murkier is that these 2 steps cannot be performed independent of each other. The authors merely address strategies for the 1st step, but it doesn’t help much in a closed-ended search, because you cannot perform the 1st step without keeping the 2nd step in mind!
For example, if you were looking for a designer and you first found creative individuals, but they all turned out to be writers, it doesn’t help you.
In a closed-ended creative talent search, you could have a creatively talented person who is not a good fit for the job at hand. You don’t want a creatively talented plumber to be doing the role of a writer. Or you don’t want a creatively talented entrepreneur to be a researcher. In fact, I don’t think you’d want Daniel and Tyler to switch jobs. While they’re both creatively talented, and would do fine if they switched roles, perhaps they wouldn’t be doing the best of what they’re capable of. Tyler is a fantastic podcaster and researcher, while Daniel is an entrepreneur and investor.
Once I grasped the distinction between open-ended and closed-ended creative talent search, I realized that applying the book's tactical advice on identifying talent to closed-ended talent search may not be suitable. However, let's take a closer look by diving into some of the specific strategies the authors mention and the questions they raised in me.
Overview of the book’s discussion on Interview Strategies
The book’s most tactical advice on identifying talent is centered on interviews. How to interview, examples of questions to ask, which interview questions to avoid, how to conduct interviews online, how to build trust in interviews etc. are some of the matters discussed.
The authors say that given how important interviews are in assessing a candidate, asking cliched questions on strengths and weaknesses, or questions for which the interviewer has prepared, doesn’t yield much into one’s creative potential. Instead, people should be asking unconventional questions such as:
- What are the open tabs on your browser right now?
- What’s the farthest you’ve ever been from another human?
- What’s something weird or unusual you did early on in life?
- How do you feel you are different from the people at your current company?
- What subreddits, blogs, or online communities do you enjoy?
- What is something esoteric you do?
The authors say that questions such as the first one on the above list hint at people’s “intellectual habits, curiosity, and what a person does in his or her spare time, all at once.”
For example, Tyler’s response to the question “What are the open tabs on your browser right now?” is “his blogging software, two email systems, Twitter, the Google Doc for writing this book, another Google Doc for another writing project, WhatsApp, his calendar, a friend’s blog, an article on quantum computing, his RSS feed, a podcast on chess, an article on the progress in the life sciences, an article on online interviewing, a French radio station [Fip] that plays Jamaican dub, and an article on in-migration to Poland.”
My Questions and Critique on Interviews: Does weirdness translate to talent?
On reading Tyler’s response, my first impression is that he is an interesting person! In fact, it strikes me that most of the questions suggested by Daniel and Tyler are trying to capture signals for how interesting or quirky one is. By posing questions such as what’s something weird one did, or what’s esoteric about them, or how far they’ve been from another human being, implicitly one is searching for what I’d call the “interesting meter” of this person. The underlying premise is that intellectual curiosity translates to creative talent, one which most of us can make some case for.
However, when does this premise fail? When do the questions suggested by Tyler and Daniel mislead one? The authors discussed why traditional bureaucratic questions fail in identifying creative talent. I’m also curious to know the limitations surrounding the tactics suggested by them, and as such I have many questions:
(i)What about interesting curious people who are not creatively talented for the job?
At first glance, it seems true that being interesting would result in creative talent - I know people who are intellectually curious and talented. However, I also know people who are interesting but were not talented at their jobs. I once worked for a startup that had a lot of interesting people – they wrote books, started businesses in their spare time, traveled the world, and had eclectic interests. But this edginess didn’t necessarily translate into talent for their job and providing value to customers. Interesting meter wasn’t a good proxy in this case. Hiring such people only based on how interesting they were without taking into account their creative talent for the specific role they were hired for was a mistake.
How often would you find people who have incredibly interesting answers to the questions posed by the authors, have climbed Mount Everest, know about Japanese music, play the guitar in their spare time etc., but are not creatively talented for the job at hand? People can be creatively talented in some dimensions of their lives, but not in others. The authors assume that a creatively talented person would be creative in every dimension.
I’m interested in knowing: When is one’s interesting meter and intellectual curiosity a good measure of talent? How good of a measure is it? When does it fail? Is it a measure of talent at all? What is the false signal one can get from these questions? Does interesting meter only work for open-ended creative talent search and not closed-ended talent search?
(ii)Are there creative talented people who lack intellectual curiosity?
What about the converse – could you have someone who is extremely talented for the job at hand, but watches trash TV in his spare time, not indicating a high level of intellectual curiosity? Biographies of creative people often amplify how curious such people were in their spare time. Is this why this idea is also magnified in our heads? Are there counter examples? I side with the authors here and am not sure if there are many counter examples, but I’d be curious to know if there are.
(iii)What about people who are so focused on work that they don’t do much in their spare time except perhaps exercise, spend time with family, read?
The authors state in the book that you want to look for people who do interesting things in their spare time, because “personality is revealed on the weekends.” People who deliberately try to practice and get better at a craft everyday show indication of talent. But what if some people are so focused on work that they don’t have many hobbies? What about people who have kids to take care of outside of work which prevents them from inculcating too many interesting side projects? Are such people not talented then? And didn’t Steve Jobs famously focus on Apple (and Pixar) all day and otherwise would sit in his room and think for several hours in his “spare time.” I don’t know if he was perusing content on subreddits and doing side projects on the weekends. Is he just an exception then? Or does talent identification based on side projects only apply to people doing jobs, not those who are entrepreneurs?
Endnotes on Interview
Daniel and Tyler point out why the conventional questions fail, and I would have loved to see some counters to their own arguments – what should the reader be mindful of when deploying these questions? Are there false signals to watch out for? And if someone does seem like a creative person based on their answers to questions, what is the next step in determining whether they are talented for the job at hand?
This brings me to my 2nd critique of the book.
Work Samples: An error of omission?
Following the advice of Tyler and Daniel would result in one deploying their suggested interview questions, followed by conducting referral interviews. If all checks out, what next? Should you jump to hiring the person?
Is there value in getting the candidate to do work trials? This is an unanswered question and I’m surprised that the book has no mention of work samples.
It strikes me that proof of work should be an important part in identifying talent, or even if it’s not, it’s worth mentioning why that’s the case. However, the book doesn’t state a word about it and as a reader I’m left empty without knowing their research and opinions on this topic.
In fact, I know that Daniel would ask for work samples when hiring. In 2019, I know someone who interviewed for a position at Pioneer (company run by Daniel) and had to do a work sample even before the interview with Daniel. So, I’m intrigued to know if he changed his mind here.
I have many questions here: What is the role of actual work samples in judging talent? Isn’t proof of work a better measure than interview answers? If not, why not? If polished speakers can give off false impressions, aren’t work samples not only harder to manipulate, but also a much better indicator of how the person thinks and produces quality work?
What is IQ measuring?
In Chapter 4 of the book, the authors explore the question “What is Intelligence good for?” They discuss whether IQ tests are a good predictor of talent, if they’re priced into the market already, and when they are useful. I think this discussion is helpful.
However, the premise of these arguments is that IQ tests are a measure of intelligence – the authors speak of intelligence and IQ interchangeably. My question to them here is: Is IQ a measure of intelligence? What does intelligence here signify? It strikes me that IQ is a measure of memory, processing speed, and the ability to do well on such tests. It also seems as if people can get better at IQ tests with practice. In that case, is doing well on IQ tests akin to skill development?
I would have loved to see more discussion on the underlying premise of 1) what IQ tests are really measuring 2) Whether IQ tests are even a good measure of intelligence 3) If IQ tests measure intelligence, what type of intelligence is it?
If creative talent is not transferable across roles, do referral interviews work?
The authors cite the importance of conducting referral interviews before hiring someone, especially for high-level jobs.
However, if creative talent depends on the role you’re doing, this implies that talent is not always transferable across unrelated roles. In this case, asking previous managers about how talented a person is may lead one astray. It’s quite possible that the person was not talented for one job but is extremely talented for another.
Scott Fitzgerald worked at an advertising company for a few months before he devoted his full attention to writing. If the publishers asked the advertising company about how creative Scott was, I don’t know if the answer would be positive. Einstein worked as a patent clerk for years while working on his scientific papers. I don’t know if his boss at the clerk office would give him a recommendation for University. Robert Greene worked 40 odd jobs before his first book assignment which was very successful. I don’t know how many of the employers in his 40 odd jobs would give him references to be an author.
Where background checks are useful is in checking for qualities like hard work, reliability, ethics. These are important learned traits that are transferable across jobs. But what if creative talent is not transferable across many jobs? Do referrals not work in such cases?
Conclusion
I think the book Talent has a lot of useful ideas. I don’t think anyone would disagree that traditional bureaucratic approaches don’t do a good job in creative talent identification. We need to experiment with novel ways of finding creative talent, and this book gives us pieces of the puzzle to help with it.
However, my main question to the authors is: Is their book addressed to only those engaging in open-ended creative talent search, which is a tiny sliver of talent search? Or do they also hope for people engaging in closed-ended creative talent search to apply their strategies? Making this distinction is important as it changes how decisions around talent identification are made. Hence, without a clear distinction, the book raised in me several unanswered questions which I’ve explored in the essay.
Find me on X @aasthajs and please subscribe for more essays.
Granted the authors do state that they view creative talent “regardless of the context”, but they also state that they write this book because traditional bureaucratic methods for identifying creative talent fail, which leads me to discern that perhaps they’re not only writing this book for those identifying creative talent regardless of context. Clarification on their open-ended view of creativity would have been helpful.