3 ms·
This is why I honestly don't care about a lot of elitism in things like schools, internships, places of work, etc. Because if anyone has the base knowledge you
by godelski 3y ago
This is why I honestly don't care about a lot of elitism in things like schools, internships, places of work, etc. Because if anyone has the base knowledge you can turn them into a top 1%er really fast and easily. But what a lot of people don't understand, is that everyone needs training. The good argument against this is that those with elite credentials have gotten training through their elite status[0] (momentum is a powerful force) but I don't find this particularly convincing because it's implying that you don't need to train people. People should always be trained. I'm not sure why we often don't internalize this. Even top athletes are constantly training, arguably that's their entire job and weirdly enough you don't even need to have the trainer be a top athlete (just sufficient knowledge of the domain and more about skills in communication and deconstructing problems). Realistically people will just settle into wherever they've found some level of success because, like the article says, they're not looking for feedback or if they are, it's not straightforward or easy to come by.
Are we coders/scientists/engineers/etc properly training? Do we have sufficient means? Are code reviews actually training exercises combined with quality control or are they an adversarial game where jobs need to be justified to those using poor metrics for said justification? (do you intentionally put in easy mistakes so review has something to point to but you have something easy to fix but would get something hard to fix if you submitted better code in the first place?)
[0] I'll put it this way, when you look at top performers in a field (let's say ML for example, because it's hip) you'll typically find a long chain of elite statuses (plenty of exceptions, but they don't discredit the underlying premise here). Take for example the well known Andrej Karpathy. Before he was at OpenAI he was the Sr Director at Tesla, and before that OpenAI, and before that Standford, then University of British Columbia, and then University of Toronto (where he took classes and attended study sessions by Hinton). The point here is not to say he didn't work hard (there's no doubt of that) or isn't exceptionally skilled (entire premise is based on assuming this), but that there's compounding opportunities to catapult one's self to the top skill level.
The truth of the matter is going into college the vast majority of people are dumb as shit and the differential in someone at a low tier school and a high tier school isn't that huge (we can even extend the argument further backwards). 4 years is not a lot of time, nor is the 8-12 for PhD (from high school). The skills gap does widen between these groups, but not nearly as much as they do between those leveraging their opportunities or even just those with better opportunities. There's a saying I like: "The harder you work, the luckier you get." The two are deeply intertwined and working hard you have the capacity to take advantage of more opportunities.
The three questions here are: 1) how many additional Kaparthy level people did we miss out on due to them not having adequate opportunities? 2) Is there a cost effective solution to increase these numbers? Dramatically? 3) At which point (for which skills) is there no longer a sufficiently effective strategy that's also cost effective. (strategy selection is most certainly time and domain dependent) There's actually decent evidence that you can take low quality high school students and they will be successful at elite schools, especially after you control for dropouts due to higher rates of family issues. (This can be captured in many ways and is nuanced so it's actually deceptively hard). In high likelihood you could randomize Kindergartners into the elite path and non-elite path and find this effective but we're probably only going to be able to get natural experiments. Is this true for High School? College? Grad School? (If you think in any way I'm undermining Karpathy's success or talent, you've most certainly misread and/or I've most certainly greatly miscommunication and this line should be a verification of the former. Karpathy is without a doubt, a fucking beast)