# Why Engineers Should Question Constraints Before Optimizing

- Original post: https://www.twitter.com/i/status/2121797534016536908
- Source account: [@bowie_HQ](https://x.com/bowie_HQ)
- Source language: English
- Duration: 2:06
- Published: 2026-11-17
- Canonical URL: https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing

## Transcript

[0:00](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing) And you, you should actually take the approach that the, the constraints that you are given are guaranteed to be some degree wrong.

[0:05](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=5) Guaranteed to be some degree wrong. Huh.

[0:07](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=7) Uh, 'cause the, the counterpoint would be that they are perfect.

[0:09](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=9) The-- Right, which is never. It's-- Um, as you say, like, what are the-- what's the probability that this is a platonic ideal of the perfect part?

[0:15](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=15) Right. Zero. Zero. Okay? Basically. Yeah. So, you question your constraints.

[0:19](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=19) It does not matter if the person handing you those constraints won a Nobel Prize.

[0:23](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=23) They are-- Even Einstein was wrong some of the time.

[0:25](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=25) Right. So, uh, question your constraints.

[0:27](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=27) This is extremely important. Um, and like an-- I think another thing that, like, if you say, like, what, what are the mistakes that smart engineers make?

[0:36](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=36) Um, like, I think that one of the most-- one of the biggest traps for smart engineers is optimizing a, a thing that shouldn't exist.

[0:46](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=46) Yeah. Um- So they'll just sit there and spin on that thing that's just like, why do we even have this in the first place?

[0:50](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=50) Well, we-- Absolutely. Yeah. So the-- because, um, when, when you go through college, you know, like studying physics and engineering, I studied physics, uh, the-- you have to answer the question that the professor gives you.

[1:02](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=62) Mm-hmm. You don't get to say, "This is the wrong question." Right, right, right. Um- Yeah.

[1:06](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=66) But in reality, uh, we have far more degrees-- When you're in reality or the-- you have all the degrees of freedom of reality.

[1:13](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=73) And so the first thing you should say is, "This question is wrong." Yeah. And that's what you said last year. Yeah.

[1:18](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=78) I mean, you kind of said like, "It took us a long time to frame the question even."

[1:21](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=81) Yes. Because we didn't necessarily know what it was.

[1:22](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=82) It took ages to frame the question. Yeah.

[1:24](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=84) It's just like The Hitchhiker's Guide to the Galaxy.

[1:26](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=86) Douglas Adams, best philosopher ever. Maybe the-- I think best book in philosophy ever, Hitchhiker's Guide to the Galaxy.

[1:32](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=92) Um, this book is so deep, uh, people don't even understand.

[1:35](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=95) Uh, but like in The Hitchhiker's Guide to the Galaxy, the Earth is a giant computer. Right.

[1:39](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=99) And, and, and the Earth, uh, it comes up with the answer 42. Okay. Yeah. Um, and, and to the question, like, so what's the answer to life, the universe, and everything?

[1:49](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=109) Answer is 42. And like, they're like, "What the hell?

[1:51](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=111) That doesn't make any sense." Right. It's like, oh, you-- The really-- The hard part is the question.

[1:55](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=115) Is the question. The answer is the easy part.

[1:57](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=117) You need a much more powerful computer to tell you what the question is. Right.

[2:00](https://www.translate.mom/transcript/xP8OaHbMg80Q/en/why-engineers-should-question-constraints-before-optimizing?t=120) And this is true. Once-- The point at which you can properly frame the question, the answer is comparatively easy. Right.
