Episode Summary
Executive Summary: Patrick O'Shaughnessy interviews Daniel Gross about identifying talent, building Pioneer, using questions and psychometrics to screen founders, and where tech investing is heading—from GPT-3 to satellites. The conversation argues that progress, persistence, and measurable feedback loops matter more than credentials or self-reports.
Main Topics: Interview questions as talent signals (Priority: 5/5): Movie tastes and simple prompts reveal work style, motivation, and authenticity. Pioneer’s mission and model (Priority: 5/5): Pioneer uses software to find, qualify, and encourage overlooked founders globally. How to measure talent and progress (Priority: 5/5): Gross favors progress-based scoring, follow-on funding, and compounding improvement over static traits. Psychometrics: useful but limited (Priority: 4/5): Psychometric tools can aid screening, but self-assessment and low effect sizes make them noisy. Insecurity, persistence, and hierarchy (Priority: 4/5): Insecurity can fuel energy, while founders often maximize the wrong status games. Frontier tech themes (Priority: 5/5): He highlights GPT-3, chat apps, social/video, and satellites as active startup frontiers. Technology becomes utility (Priority: 5/5): Consumer tech often evolves into regulated infrastructure, changing how investors should think about it.
Key Arguments: Whiplash screens for work-as-pursuit vs work-as-escape. Great founders need many shots on goal and visible progress. Follow-on funding helps, but alone can't prove talent. Psychometrics help with taxonomy, not truth; self-assessment is noisy. Insecurity can create energy, but can also drive bad status maximization. Many tech firms are becoming utilities with regulatory lock-in. GPT-3 is compelling because it’s a visible new capability with real demos.
Data Points: Follow-on funding rate at Pioneer: 20, 25% - Share of funded founders later funded by others Effect size mentioned for IQ to earnings: 0.6 - Used as an example of psychometric predictive power Version of Pioneer’s current talent view: 16-bit bitmap - Metaphor for low-resolution current scoring Future Pioneer resolution target: 4K edition - Metaphor for a much richer metric system Current Pioneer scoring inputs: week over week - Score improves as founders make measurable progress Startup validation goal: 10 - First milestone for a product to attract avid users Second growth milestone: 100 - Next target after reaching 10 users Revenue milestone: $1,000 in recurring revenue - Step after reaching 100 users Typical GPT-3 output quality: 6 times out of 10 are bad - Gross describing model reliability OpenAI model creation cost: single digit millions - Estimate for GPT-3 training cost Historical benchmark for satellite launch: $50 million - Approximate cost of an Iridium satellite launch Projected SpaceX satellite launch cost: $50,000 - Illustrative lower-cost launch estimate Kindness story road conditions: 95 degrees - Gross describes being stuck in hot California conditions Distance run before help: 12 miles, 12, 14 miles - He returned dehydrated after a long run
Pivotal Quotes: "the art of asking great questions" — Patrick O'Shaughnessy: Describing the episode’s focus on interviews and screening "I think the question of talent to actually be the most important question in the world" — Daniel Gross: Why he studies founders and psychometrics "You don't prove what you don't measure." — Daniel Gross: On why feedback loops and scoring matter
Implications: The open question is which new ventures can turn noisy signals into durable advantages; listeners should favor measurable progress, not story or status.
From the Transcript
Patrick O'Shaughnessy is the CEO of O'Shaughnessy Asset Management. All opinions expressed by Patrick and podcast guests are solely their own opinions and do not reflect the opinion of O'Shaughnessy Asset Management. This podcast is for informational purposes only and should not be relied upon as a basis for investment decisions. Clients of O'Shaughnessy Asset Management may maintain positions in the securities discussed in this podcast. My guest today is Daniel Gross. Daniel is the founder of Pioneer, an extremely unique company which he describes as a fully remote startup generator that helps talented people around the world figure out if their idea has legs. You can learn more about it at pioneer.app. Our wide-ranging conversation covers the art of asking great questions, the use of predictive modeling and psychometrics to identify talent, and why psychometrics are probably overrated and not that scientific. We then dive into exciting new frontiers for tech investing ranging from GPT-3 to satellites. I really enjoyed this conversation and hope you will too.
I've had the pleasure over the course of my career, both through Pioneer and through others, to invest in a bunch of different companies. I've had the chance to meet a bunch of different founders and over the course of my time, having maybe been one, and I'm very interested in this topic. I think there's an interesting question for any asset allocator: to what extent do you believe, especially at the early stage, that you want to be looking for great markets or great founders? And like everything, the truth is somewhere in between. We can accept maybe as an axiom that being able to screen people, either using software or using your own mind, is an important thing, not just for investors, of course, but for Anyone looking to hire or really do anything significant in life, I find the question of talent to actually be the most important question in the world because it is, at the end of the day, required to do anything interesting. As somebody that's done a fair amount of predictive modeling, I know from experience that often the most important thing is defining the outcome very cleanly and very well before testing certain variables or features against that outcome. I'd love to hear how you define talent. If the idea is to search for talented people, how do you know that you found one in some sort of
Not just quantify who's good and who's not, but also motivate people to get better. I think if we're able to build a reliable metric for day-to-day progress, it will not only be a great service for creating more startups, but I think it'll be a great thing for productivity if you don't prove what you don't measure. And I think part of the issue is, by the way, not just for Pioneer, but I think for anyone working on something early stage or an investor, it's not really clear at the end of the day what separates a good day from a bad day. There's no score. And we know that in markets and industries where there is a score, people do improve. And markets are efficient. So, for example, in sales, it's very clear what a good day is, it's very clear what a bad day is, and people are motivated to improve. You can measure who's great and who's not. And if there's someone junior on the team who's really good, they're rewarded based on the merits of their work. In sports, this is extremely clear as well: a good day and a bad day. It's not really clear if you're an investor what a good day or bad day is. It's not really clear if you're an early-stage startup what a good day or bad day is. I mean, even for a software engineer, it's not really clear what a good day or bad day is. I'm curious to what degree still psychometrics, and maybe you can define what that term means. What are some examples?
About Invest Like the Best with Patrick O'Shaughnessy
Conversations with the best investors and business builders in the world.
View all episodes from Invest Like the Best with Patrick O'Shaughnessy