logoalt Hacker News

incrudible • today at 12:06 AM • 1 reply • view on HN

> The fact that LLMs are trained on a corpus does not mean their output represents the median skill level of the corpus.

It does, by default. Try asking ChatGPT to implement quicksort in JavaScript, the result will be dogshit. Of course it can do better if you guide it, but that implies you recognize dogshit, or at least that you use some sort of prompting technique that will veer it off the beaten path.

Replies

keeda • today at 3:00 AM

I asked the free version of ChatGPT to implement quicksort in JS. I can't really see much wrong with it, but maybe I'm missing something? (Ugh, I just can't get HN to format code right... pastebin here: https://pastebin.com/tjaibW1x)

----

function quickSortInPlace(arr, left = 0, right = arr.length - 1) { if (left < right) { const pivotIndex = partition(arr, left, right); quickSortInPlace(arr, left, pivotIndex - 1); quickSortInPlace(arr, pivotIndex + 1, right); } return arr; }

function partition(arr, left, right) { const pivot = arr[right]; let i = left;

  for (let j = left; j < right; j++) {
    if (arr[j] < pivot) {
      [arr[i], arr[j]] = [arr[j], arr[i]];
      i++;
    }
  }

  [arr[i], arr[right]] = [arr[right], arr[i]]; // Move pivot into place
  return i;
}
➕ show 1 reply