liftsh

5 ms·

1.
by liftsh 3y ago
A notoriously hard class and one of my favorite ones. Stefano Ermon was a great instructor.
2.
by liftsh 3y ago
To the contrary, the tools already make a material difference. I wrote 10k lines of code in the last 3 weeks--maybe 60% of that was generated by ChatGPT. Sure, I could have done it myself, but if it wasn't for AI my hands would have li
3.
by liftsh 3y ago
This is a good perspective, but I'm not sure I fully agree. I wonder if it's just that the nature of work, the nature of "what pays well," has changed. Sure, my grandparents had a nice house, but they also worked the min
4.
by liftsh 3y ago
Yeah, good point--when I said new wasn't always better, I was just talking about the case where the new tech solves a problem, but it's not the one you have. Like choosing GraphQL just because it's new, even if your data does
5.
by liftsh 3y ago
Reminds me of Resources from Scala. I keep seeing features being carried over from language to language--I wonder if we'll eventually reach convergence where all languages, perhaps within several broad groups like OOP vs FP, look the s
6.
by liftsh 3y ago
And the thing that links those two groups of people is their desire for things. Never was religious, but this core point of Buddhism (from an ELI5, admittedly) really resonated with me. Wanting fewer things indeed seems to makes you happier
7.
by liftsh 3y ago
Expectations are important and are missing from the equation here. Not everyone needs children, a fat pension or a large house to be fulfilled. It is also perfectly fine to be with the same group of people, have the same routines and the sa
8.
by liftsh 3y ago
Hmm...is this really surprising to anyone? Yes, I'd like the purpose of a tech company to be to "make developers heroes," but in a capitalist society, companies exist to make money. I'm not saying I like it, but that&#x2
9.
by liftsh 3y ago
New is not always better, but many times it is. We see this for example in programming languages, where newer ones incorporate the best features of their predecessors. I think there are two things to be wary of: 1) Selecting a new technolog