Lambda Coding Boot Camp Claims 86 Percent of Students Get Tech Jobs. They Don't.
If you visit the website of Lambda School, a "boot camp" for people who want to quickly learn how to code, you're…
Vincent Woo writes a scathing piece about Lambda School in the New Yorker, but I think his perspective is off.
Disclosures first: I don’t know Austin, I have no stake in Lambda, have not had any interaction with any student, graduate, nor employee of Lambda, and am about far removed from it as possible. I do, however, have an interest in ISAs, but as of yet no professional ties to any organization pushing ISAs yet.
Assorted thoughts on the article:
- My own experience with bootcamp graduates is pretty dim, but I am still optimistic that overall the bootcamp industry can figure it out. Looking at my own software engineering career, I would say I consistently use about 3 semesters worth of material. I don’t expect anything less than half a year of very rigorous training to come anywhere close to matching a new graduate in skill, and even then it’s a tough proposition.
- A 50% job placement rate for a bootcamp isn’t bad. The article is mostly predicated on Lambda’s misleading claims and disorganized curriculum, but it doesn’t seem to indicate any graduates are laden with debt in the same way 4 year liberal arts graduates might be.
- Financing an ISA vs “selling,” which the author says is a “meaningless semantic distinction” is most certainly not a meaningless semantic distinction. When selling a contract, Lambda is off the hook for the performance of the student, ie whether or not the student gets a job doesn’t matter in the short term. When taking a loan backed by the ISA, Lambda still forks up interest payments, even if the student doesn’t get a job. In the short term, a loan backed by an ISA requires more skin in the game than not.
- PE bought Lambda ISA contracts at $10k per contract, not total. This was corrected by Vincent on Twitter, but the wording in the article is unintentionally misleading.
- Lambda ISAs, as of this writing, expire after 5 years. That’s way too long. In a market in which software engineers are a scarcity, a program that claims to train people effectively should be confident in its ability to place workers within half a year or year.
I get it. Being mislead on the effectiveness of a program really stinks. But the cold hard truth is that software engineering is very very difficult, and while I think the vast majority of people can become reasonably good engineers, there’s an enormous amount of variance in the required time to get up to speed.