3 ms·

I can’t fathom a world where an LLM would be able to review code in any meaningful way -at all-. It should not substitute a human, and probably wasted more eff

I can’t fathom a world where an LLM would be able to review code in any meaningful way -at all-. It should not substitute a human, and probably wasted more effort than it solves by a wide margin.

hsbauauvhabzb 2y ago
And for any poor soul forced into an environment using AI driven commit reviews, be sure you reply to the commit reviews with gpt output and continue the conversation, eventually bring in some of the AI ‘engineers’ to help resolve the issue, waste their time back.
dakshgupta 2y ago
I understand your skepticism and discomfort, and also agree that an LLM should not replace a human code reviewer. I would encourage you to try one (nearly all including ours have a free trial). When done right, they serve as a solid first pass and surface things that warrant a second look. Repeated code where there should be an abstraction, inconsistent patterns from other similar code elsewhere in the codebase, etc. Things that linters can’t do. Would you not want your coworkers to have AI look at their PRs, have them address the relevant comments, and then pass it to you for review?
- smallpipe 2y ago
  > Would you not want your coworkers to have AI look at their PRs, have them address the relevant comments God no. Review is where I teach more junior people about the code base. And where I learn from the more senior people. Either of us spending time making the AI happy just to be told we’re solving the wrong problem is a ridiculous waste of time.
- hsbauauvhabzb 2y ago
  No, just like I don’t want coworkers using LLMs to generate code. LLMs have their uses, producing quality code is not one of them, they generate cancerous outcomes.
makingstuffs 2y ago
Yeah it seems like the inverse of what is logical: Have a human review the code which _may_ have been written/influenced by an LLM. Having an LLM do code review seems about as useful as YOLO merging to main and praying to the Flying Spaghetti Monster...

I can’t fathom a world where an LLM would be able to review code in any meaningful way -at all-. It should not substitute a human, and probably wasted more effort than it solves by a wide margin.

hsbauauvhabzb 2y ago

And for any poor soul forced into an environment using AI driven commit reviews, be sure you reply to the commit reviews with gpt output and continue the conversation, eventually bring in some of the AI ‘engineers’ to help resolve the issue, waste their time back.

dakshgupta 2y ago

I understand your skepticism and discomfort, and also agree that an LLM should not replace a human code reviewer. I would encourage you to try one (nearly all including ours have a free trial). When done right, they serve as a solid first pass and surface things that warrant a second look. Repeated code where there should be an abstraction, inconsistent patterns from other similar code elsewhere in the codebase, etc. Things that linters can’t do. Would you not want your coworkers to have AI look at their PRs, have them address the relevant comments, and then pass it to you for review?

smallpipe 2y ago
> Would you not want your coworkers to have AI look at their PRs, have them address the relevant comments God no. Review is where I teach more junior people about the code base. And where I learn from the more senior people. Either of us spending time making the AI happy just to be told we’re solving the wrong problem is a ridiculous waste of time.
hsbauauvhabzb 2y ago
No, just like I don’t want coworkers using LLMs to generate code. LLMs have their uses, producing quality code is not one of them, they generate cancerous outcomes.

makingstuffs 2y ago

Yeah it seems like the inverse of what is logical: Have a human review the code which _may_ have been written/influenced by an LLM. Having an LLM do code review seems about as useful as YOLO merging to main and praying to the Flying Spaghetti Monster...