Talent Search versus Talent Development (2019) [pdf]
community.ams.org
community.ams.org
We had to let them go. Then we hired an experienced dev who was self driven and the difference is night and day. The senior dev costs slightly more than the junior dev.
I don't think I will ever try the "talent development" approach again. The talent needs to do that work itself. If it is unwilling or unable, then so am I.
FWIW, I'm currently on this trajectory (as the mentor).
Activation energy cannot be trained. And that’s the prerequisite for just about everything, especially propensity for improvement.
We cannot develop talent in the same way pro sports teams do because there’s not a 50x difference between what we pay and junior or senior dev like there is in the NBA or a 1000x difference like in MLB or Euro soccer
The economics don’t support it. That’s the secret reason why schools matter still.
It always baffles me how many places only pay sr engineers 3-5x more.
I have worked with those people and most of what they created was a heap of code that someone had to maintain after they left.
It's not hard for one or two decently talented engineers to accomplish more in a year than 50 people.
But I have the complete opposite experience. I'm a junior dev working on an embedded linux system, and some associated test applications. I have an extremely good mentor/boss who's helping me develop skills for those environments, and other associated skills that are absolutely invaluable to my career.
Now the difference of course is I am highly motivated. I love the work I do, and because it's so enjoyable and fun, I absolutely take a lot of my personal time to play around with work-related or at least work-adjacent projects.
My worry is that when I do eventually move to my next company/product, Ill be assigned to a manager that has had too many poor experiences with unmotivated juniors, and is unwilling to share their experience with me (in a cooperative manner, not just a one-sided relationship of course).
Talent does not grow on tree on its own but in organizations like yours, but not everyone is destined to be a talent. Those without the proper attitude for becoming a talent had to be let go, so the outcome is the same, only arguing the conclusion. You had bad luck with 3 bad ones for one good, but the usual ratio is not too far from this actually for other organizations. When you find the good attitude people it is a joy, especially if they grow into your environment, being tailor made for your figure. That will even be better than the one reused from elsewhere. ; )
Not that you can expect everyone to learn everything, or jump to a 'higher' level merely by having access the right information.
Put them in a situation where they need to produce a result, give them support, but if they need hand holding give them that feedback: you need to be independent on this or you will not be retained.
It sucks, but one or two experiences writing code for a junior dev and then seeing they do not even move on from there was enough. You can have the same kind of issues with crappy “senior” colleagues too, though…
Of course, if you don’t have the right kind of task for an “advanced beginner” (so to speak), that is a problem on the management end.
Bringing up junior colleagues is a fraught business, but when it succeeds you get a LOT of results out of it.
In the end, I try to keep in mind that it is all about money and time. From an individual project point of view, though, it can be very satisfying.
I think the problem is that it's really hard to identify people worth taking a risk on developing, because during interviews everybody puts their best foot forward. Typically people use some kind of proxy for this like education or prior achievements, but those are far from perfect (and in fact, both are heavily gamed such that they are rapidly declining in value as a signal) because people will have varying levels of motivation throughout their life and in different environments, and people lie/exaggerate. You can also filter on how much they've developed themselves/been developed, eg how well they perform in a coding interview or what kind of personal projects they've done, but that's arguably just filtering on experience and also subject to gaming (people do these things just to get a high paying SWE job and that drive isn't necessarily there after they have one).
Also, you have to be willing to let people fail or struggle with tasks (you think are/that are) beyond their current skillset to let them grow. To a certain extent people can be trained to persevere through or creatively work around problems, although I do think there is a lot of variance in how well people pick that up. I've found that being too willing to help with these problems is a major reason people don't develop those skills themselves - it's just hard because they're often a source of frustration or stress for the other person, and you know you can probably solve it, and they might know you can probably solve it, so you have to be the bad guy and make them do it anyway.
I recently made the first hires at my startup and handled everything from the job description to talent sourcing to interviewing to onboarding, looking at early career people. Because the hiring market is so favorable to employers right now, I got a ton of applications and managed to hire some really great people. One thing I noticed is that the best candidates tended to be "unique" in their resumes or even their way of contacting me about the position, and they also had strong interests such that they had deliberately gone out of their way to learn new things beyond trivial stuff like "side projects" that amount to following a tutorial. Sometimes a resume might suggest that, but in an interview you could see that those interests were purely surface level; I also designed a pair-programming interview designed to put them way out of their comfort zone as a way to evaluate their perseverance and creativty. So far the agency and curiosity my hires demonstrated has extended into their jobs, and I'm really happy to be able to develop them further, because I can trust that they're going to be active participants and not just "some person who wants a job" that expects to be spoonfed.
When I got my start, I was the only engineer in a call center location. About 5 people a year would ask how they could do what I do.
What I learned after writing a long essay, with YouTube links and book references even, is that people just like the idea of it, without the work part. Not one did any of what I laid out.
After that, anyone who asked I just mumbled something about college, even though I never attended one.
I think this is why YC sends notes to rejected applicants who are in "the top 10%" by whatever standard they are grading folks on, and encourage them to apply again.
Talent _development_ requires substantive feedback, or at least substantial visibility into the competition versus a binary outcome. For example, tell Founders if they have the substance or right fidelity but poor sense of timing. (And market timing often in the hands of the VCs anyways).
Without real feedback, it’s much closer to a real estate license than an academic math program.
All that said, my impression is current math clubs and competitions have been taken over by kids just wanting credit for college rather than the kids of my era, who were there 'cause they actually enjoyed math.
If the only purpose you can see for math competitions is find a few very talented kids, yeah, well then they would qualify as counter-productive.
I work directly with students in this field and my experience is that most are involved for the interest, challenge, or social aspect.
If you are involved with “pay to play” programs like AoPS, you probably see more resume-grooming.
Nobody I know would continue to donate their time if they perceived it was primarily fluff.
As far as I can tell, this has only happened once, in 1939 [1].
[1]: https://www.snopes.com/fact-check/the-unsolvable-math-proble...
There are paid online courses as well as free purses run by students, some of whom are trying to polish their resumes for college apps.
We are in a golden age for student mathematics study.
What's missing is a strong environment for youth mathematical research. There is a collection of summer camps, but they are too few and expensive.
I think the bigger point is that in a functioning economy, companies would not have any choice but to train in house, because skilled labour looking for work would not be found in excess.