3 ms·

That "a good model (if you know how to operate it well)" is doing a lot of lifting. To be sure, there are a lot of bad books, and you can get negative advice f

by saulpw 1y ago

That "a good model (if you know how to operate it well)" is doing a lot of lifting. To be sure, there are a lot of bad books, and you can get negative advice from them, but a book has fixed content that can gain and lose a reputation, whereas a model (even a good one!) has highly variable content dependent on "if you know how to operate it well". So when someone or some group that I respect recommends a book, I can read the words with some amount of trust that the content is valuable. When someone quotes a model's response without any commentary or affirmation, it does not inspire any trust that the content is valuable. It just indicates that the person has abdicated their thought process.

simonw 1y ago
I agree that quoting a model's answer to someone else is bad form - you can get a model to say ANYTHING if you prompt it to, so a screenshot of a ChatGPT conversation to try and prove a point is meaningless slop. I find models vastly more useful than most technical books in my own work because I know how to feed in the right context and then ask them the right questions about it. There isn't a book on earth that could answer the question "which remaining parts of my codebase still use the .permission_allowed() method and what edge-cases do they have that would prevent them from being upgraded to the new .allowed() mechanism"?
- theamk 1y ago
  And as long as you don't copy-paste its advice into comments, that's fine. No one really cares how you found all those .permission_allowed() calls to replace - was it grep, or intense staring, or AI model. All that matters is you stand behind it, and act as an author. Original post said it very well: > ChatGPT isn’t on the team. It won’t be in the post-mortem when things break. It won’t get paged at 2 AM. It doesn’t understand the specific constraints, tech debt, or your business context. It doesn’t have skin in the game. You do.
  - tensility 1y ago
    Further, grep (and any of its similar siblings) works just fine for such a task, is deterministic, won't feed you bullshit, and doesn't charge you tokens to do a worse job than existing free tools will do well. Better yet, from my experience with the dithering pace of LLMs, you'll get your answer quicker, too.
- 000ooo000 1y ago
  >There isn't a book on earth that could answer the question "which remaining parts of my codebase still use the .permission_allowed() method and what edge-cases do they have that would prevent them from being upgraded to the new .allowed() mechanism"? You're so close to realising why the book counter argument doesn't make any sense!

by saulpw 1y ago

simonw 1y ago

I agree that quoting a model's answer to someone else is bad form - you can get a model to say ANYTHING if you prompt it to, so a screenshot of a ChatGPT conversation to try and prove a point is meaningless slop. I find models vastly more useful than most technical books in my own work because I know how to feed in the right context and then ask them the right questions about it. There isn't a book on earth that could answer the question "which remaining parts of my codebase still use the .permission_allowed() method and what edge-cases do they have that would prevent them from being upgraded to the new .allowed() mechanism"?

theamk 1y ago
And as long as you don't copy-paste its advice into comments, that's fine. No one really cares how you found all those .permission_allowed() calls to replace - was it grep, or intense staring, or AI model. All that matters is you stand behind it, and act as an author. Original post said it very well: > ChatGPT isn’t on the team. It won’t be in the post-mortem when things break. It won’t get paged at 2 AM. It doesn’t understand the specific constraints, tech debt, or your business context. It doesn’t have skin in the game. You do.
- tensility 1y ago
  Further, grep (and any of its similar siblings) works just fine for such a task, is deterministic, won't feed you bullshit, and doesn't charge you tokens to do a worse job than existing free tools will do well. Better yet, from my experience with the dithering pace of LLMs, you'll get your answer quicker, too.
000ooo000 1y ago
>There isn't a book on earth that could answer the question "which remaining parts of my codebase still use the .permission_allowed() method and what edge-cases do they have that would prevent them from being upgraded to the new .allowed() mechanism"? You're so close to realising why the book counter argument doesn't make any sense!