Comment by godelski
2 days ago
> Why not?
Weird egos. I moved from academia to industry and constantly got told "In industry we just care that 'it works'". I thought that was a weird premise, given... you know... who doesn't? But the more time I spent in industry the more time I found that they in fact do not care if it actually works. What seems to matter more is the politics and about "working"[0] the right way using the right new buzzword[1]
Truth is that the work and complexity is not that divorced. Honestly, the work in academia felt harder, though more fulfilling. Industry work hasn't made me have to really think deeply. If anything, I've heard most of my coworkers (at multiple companies) say something along the lines of "we have to move so fast that there's no time to think." Given that (multiple) managers tell me I'm "too slow" just because I'm not producing tons of lines of code (I'm neck and neck with everyone on milestones), I understand what they're talking about. Industry has a working mode of "do first, think second" while academia often thinks first. The reason is really because it is a lot cheaper to think first.
[0] It works enough for some demo to some person
[1] One example is I beat a company's fancy giant transformer based image detector with a scrappy CNN that took only a few hours to train. They were excited for all of 1 day and then wouldn't let me do the same thing to the transformer model (which would have had a bigger impact). Fun fact, my boss also loved to tell me about how dumb academia is because they never do anything useful and how industry makes all the real advancements.
It has always seemed clear to me that the world requires two types of people.
1. "Thinkers", those extraordinarily brilliant people who spend lots of time on a single problem, discovering the optimal and most performance theoretical solution
2. "Hackers", who assemble tools by implementing designs from "Thinkers"
Without Thinkers the Hacker could not possibly solve all required problems well a single tool/product requires
Without Hackers the Thinkers work would languish in the ivory tower it was conceived in
>If anything, I've heard most of my coworkers (at multiple companies) say something along the lines of "we have to move so fast that there's no time to think.
Anecdotally, the tide on this is changing—all the low-hanging fruit has been picked and I see/hear many of my older colleagues regret not having gone the whole way with a PhD. This on top of AI being able to answer many of the engineering questions you might have learned the answers to in a masters/bachelors.
>Fun fact, my boss also loved to tell me about how dumb academia is because they never do anything useful and how industry makes all the real advancements.
He's not entirely wrong though. Industry makes the advancements that actually supposed to sell and be profitable on the free market. Academia is all over the place, as not everything being researched there can be used commercially, often it's just to get grant money, push papers and raise their egos amongst their peers.
That's not the pipeline at all. Academia creates the foundation industry sits on. There's a lot more failure, yes, but hits are way more impactful. Innovations generate entire industries.
So weird argument. Academia isn't meant to be "profitable" because no one is measuring the indirect profits. But when you do it's comically large
>Academia creates the foundation industry sits on.
Depends on the industry. All the researchers I know in academia are just wasting government grant money not delivering anything useful. Their words, not mine.
Some is useful shure, a lot is bullshit though.
Why are you conflating and equating "useful" with "sells for profit on the free market"?
There are many areas of research where profit is not a goal, and cannot be one. Understanding how and why climate changes is extremely important and useful, but cannot turn a profit. Researching different education methods, same. Hell, the researchers who won the 2024 Nobel prize in economics, who helped us understand how to build economically successful nations, something incredibly useful, cannot turn a profit with their research.
It's frankly absurd to expect everything useful to be profitable.
This conflation between "useful" and "profitable" is maybe the most annoying motte and bailey that industry people like to employ.
1 reply →
> Understanding how and why climate changes is extremely important and useful, but cannot turn a profit.
That definitely is for profit. They aren't researching climate change for the love of the game, but because agriculture, oil futures, real estate development, insurance policies, all depend on predicting climate developments.
1 reply →