D

@PTOOP @TaliaRinger @ShriramKMurthi @JAldrichPL To some extent, I think this reflects the way jobs: a second class, lower paid, more contingent position is not going to lead to serious attention. The positions are a creation of the modern university though, so not totally surprising…

Another weird thing about academia is how many things are given to people only once they no longer need them nearly as much, like extra teaching releases for tenured faculty. Wouldn't it make more sense to grant more teaching releases to pre-tenure faculty? lol

@Blaisorblade @jonmsterling Yes, and it’s pretty established that scamming is more successful if you make it ridiculous (intentional typos, classically), to weed out people unlikely to fall for rest of scam.

@jonmsterling @dbp__ Even the actually rich people do some stupid impulse buys, so why not? And "try stupid scams and see what sticks" must make *some* sense

@jonmsterling Lol yeah. Impulse “this gonna make money, I’ll figure out the rest later” buy

@krismicinski @d_feldman THE CAMEL'S HANDS ARE CLEAN!

@krismicinski I wonder if you’ve built up a much more sophisticated ad profile, so the promoted stuff you get is… relevant (ish). Whereas the default stuff that gets pushed out is just awful (I think that’s generally true of, e.g., YouTube)

@ShriramKMurthi @btbytes Says the person that’s the only reason I wrote Java in undergrad :P

@btbytes OTOH, that's the only thing that might get me to write Java.

Imagine writing Java after 4 redbulls.

@hillelogram Another difference: software engineers should know how to work on teams/projects much larger than they can tackle.

Programmer is a much broader category. Most scientists are programmers but not software engineers. One difference: I'd *encourage* a programmer to use version control. I'd *expect* a software engineer to.

@djg98115 @samth @MattZeitlin It’s not usually a ingredient we add on its own, rather than as part of@an organic compound (I guess cooking in cast iron almost counts…)

@samth @MattZeitlin We eat lots of minerals like iron, no?

This made me think, is salt the only non-organic thing we regularly eat? Maybe I’m missing something obvious but nothing else comes to mind

@EvanMPeck Yes, though I think the path is somewhat roundabout: teach about specifications, tests, and that we use that to filter out generated code that doesn't make sense. (Caveat: I'm working on the former, _not_ trying to incorporate copilot into it! Though I'd like to in the future)

Each time Github Copilot comes up in a CS ed context, it's treated as adversarial to our educational mission (I've done this too). Is anyone out there thinking about how to teach students the best way to program *with* Copilot? [1/3]

@rickasaurus @satnam6502 Sure. But classes like software dev / eng exist! Possibly even required

@dbp__ @satnam6502 The problem I see is you only get so many classes. Students have to pick and choose.

The weird thing about going into management, after 5 years I can still read and write code, but god help me if I need to do anything with the tools or build system.

@satnam6502 @rickasaurus I think there’s a lot to be said for what’s in this thread — we should absolutely be teaching the essence of modern systems, but ideally not the details (which change quickly and are easy to look up). How to do that is tricky!

@rickasaurus Knowing how to code seems to be 10% or less of what you need to know to develop and deploy modern systems. Should more of this be taught in universities? Or should university focus on fundamental concepts and principles of just programming in a specific language?

@d11a_24 @johncollins60 @moyix You’re missing the point (which I just said: maybe reread?). There’s no question that this site could be built with many fewer people. But given that it _wasn’t_, a very different question whether it can transition without significant downtime. Code reflects human organization.

@dbp__ @johncollins60 @moyix All of social media tech is/was overstaffed, not because it requires thousands of people to run an app, but b/c it's a competitive market. Competitors come up with new feat all the time that have to be implement, otherwise your gone. A nuke power plant operates with 500- 800 ppl

I have to admit I kind of like moments like this, where two wildly diverging views of the world with verifiable predictions are put forth ("Twitter will be offline within a week" vs "Business as usual"); it's a great chance to figure out who knows what they're talking about!

@johncollins60 @moyix Building something from scratch to require a small number of people to maintain (e.g., early WhatsApp) is pretty diff. from a massive distributed system that’s been built/evolved/maintained by thousands suddenly losing nearly everyone who knew how it works.

@moyix Folks might want to recall the Craigslist operates with 50 employees. Not engineers, employees.

@wilbowma @johnregehr What rock have you been under and can I come live there too?

@jonmsterling Somewhat self fulfilling though. It takes time & resources & ideas to fix such broken things… pushing it off to lower status overworked instructional staff is unlikely to ever improve things!

@jonmsterling Universities are ideally both places of research and learning, and faculty should be participating in both of those things. Now in reality universities are corrupt money making endeavors, but ideally _faculty_ push back against that…

@jonmsterling That makes sense. I think my original comment was pushback on the cultural idea among faculty (not necessarily those on here, but in general) that research is the only thing that matters, and if all teaching (& service) could go away, life would be perfect. I think that’s harmful