20 comments
Very first bullet
> AI can now generate code faster, and often better, than humans.
The author doesn't know what they're talking about. Feel free to skip this one.
[flagged]
They are certainly correct if you replace "humans" with "the average GS coder".
Devil’s advocate: “than humans”, not professional programmers
Goldman Sachs does not look for coders who have studied philosophy. For the simple reason that Goldman Sachs is not a person, but a firm. It has 16000 engineers, which means many, many different groups. The "strats" are about 25% of that, and they are split into numerous groups too. Groups are made of people, and individual people have their own interviewing style. There might be a few people who are looking for coders who have studied philosophy, but it's not a general trend. Goldman actually does have some interviewing guidelines, you need to pass some internal training to even be allowed to interview candidates. And none of these guidelines mentions anything about philosophy. They are the typical HR interviewing guidelines that you can find anywhere: technical competence, problem solving, business acumen, ethical conduct, communication, initiative, collaboration, goes the extra mile, etc, etc.
Do these people really think writing code is the hard part? It’s knowing what to write, and AI may be a multiplier there, but that knowledge isn’t obtained from asking AI the right questions.
Writing like coding requires logic/philosphical structure (not just from/part of programming language). Logically / philosophically strutured/thought out design tends toward better coding/programming results (have points of reference). At least in western world universities; logic/philosophy is what other disciplince (social, art, science & engineer) are off shoots of.
> philosophy is what other disciplines (social, art, science & engineer) are off shoots of
Yeah, but the dark side of that is that in the modern world, "philosophy" is what is left after everything useful branched off into separate disciplines.
It’s unfortunate that disciplines forget their roots in philosophy after branching out. For example, concepts such as “objective reality” or “absolute existence” are beyond the purview of natural sciences that deal in observable phenomena and models, yet people who grew up physics-first and right a good philosophical foundation tend to just believe that entities featured in those models (which, as philosophy would tell us, are all necessarily and to unknown degree incomplete and/or incorrect at any given point in time) is somehow the reference point and ground truth.
Is there really a relevant causation discussion on "oct 31 = dec 25" influence on Tim Burton writing "Nightmare before Christmas" without Tim Burton commentary?
Does seem to incorporate both the P and NP parts in P=NP; but having a directors cut doesn't prove it.[1]
[1] : https://www.imdb.com/title/tt0107688/alternateversions/
counter point:
Boolean logic/heart of computation philosophy isn't what's left over after higher abstraction levels have evolved from binary -- still need associated 'rules/terms' ways to impliment/make use of logic & rules/terms.
Just provides common reference/bridge between things that have abstracted rules/terms and implimentation/usage under different disciplines.
Would hazzard to guess idea of requiring classical concept of philosopy would be to have common reference point(s) between engineering & non-engineering disciplines.
> It’s knowing what to write, and AI may be a multiplier there
unless you think neuralink or similar is replacing your hands in a hurry, your wpm won’t be able to compete with tpm for long, no matter how fast you type, no matter where you run your models.
> , but that knowledge isn’t obtained from asking AI the right questions.
not for someone that already has it. nor for someone that lacks the words to ask.
Amazing accidental perspective on how little the typical business person or journalist understands about programming and gen ai. You can see how they think the hard part of programming is writing code, which, of course, is false. The hard part is writing the correct code and maintaining it.
I can’t wait until prompt engineering as a profession completely dies out. Its total vapor ware and its existence promotes problematic ideas about what AI is capable of
This should be backed up by GS hiring statistics only.
Well, my undergrad in CS required I take 2 classes in Philosophy. I never have directly used the material studied in those classes.
I’d note the prior CTO of Goldman was a PhD in philosophy drop out.
Nice, I'm a coder and I have studied philosophy at university!
"Goldman Sachs: When we say that morality in finance is optional, we know what we're talking about."