> As a product, Screen was wildly successful (granted, it helps to not be charging anything). As early as 2015, I just screenshotted the questions from Triple

> As a product, Screen was wildly successful (granted, it helps to not be charging anything). As early as 2015, I just screenshotted the questions from TripleByte and gave them to people to take. > The one where companies came to us (particularly post-pivot) to hire the candidates (senior engineers in the US, especially ones from top schools or with FAANG-like experience) that are hardest to acquire? Well, it turns out that companies also don't feel much need to pre-screen such candidates and want to avoid putting barriers (like, say, an online quiz) in their way. I used my ad-hoc TripleByte screen on a lot of top schools, Harvard & MIT students. Using the quizzes revealed to me that the average MIT or Harvard CS student was like 4 years ahead of knowledge in same-class Berkeley student. Way, way bigger than I ever expected. This is the real reason the Screen product failed and recruiting is hard: programming skills are not normally distributed, they are exponentially distributed. If you didn't believe in the 10x engineer, you better start believing now. If you had worked as a recruiter, you'd know the most successful ones do not "just" source better candidates, which you have been belaboring is impossible for a while. They either have tremendous volume, or they are worm-tongued with hiring managers: while skills are exponentially distributed, an average CTO can be persuaded by a very basic, LA Central Casting-level of charisma and good looks.