3 ms·

While I also view this development critically, why do you assume AI will be unable to fix the issues eventually?

• yoyohello13 6mo ago
  Whether they can or not in the future is kind of irrelevant. The fact is that right now they are not able to, but many are using them as if they are.
  • mycall 6mo ago
    I tend to disagree. When guiding AI through many rounds of code review, it can self correct if shown where general issues exist. It does take practice for using the language of the model, i.e. drift instead of issues. Human in the loop is good enough to produce useful and accurate code today.
    • shimman 6mo ago
      If you can actually do this, please sell your services. You will become a multi-millionaire overnight if you can provide a workflow that doesn't result in mass hallucinations or incorrect suggestions you're able to do something no other LLM company can. The more common use case is that these tools struggle immensely on anything outside the happy path.