18 ms·

I don't understand why people insist in giving instructions to a chatbot when we know it works by generating texts based on prompts.

wizofaus 3y ago
You might as well ask why people bother asking it questions or talking to it at all on that basis. The fact it does typically respond differently to commands than to questions is part of the power of how LLMs work (whether that's entirely due to the fact its training data includes examples of commands being given and appropriate responses or it's explicitly part of the programming I couldn't venture to say, though I suspect the former).
- namaria 3y ago
  Talking to it... I get. It is a chat bot. Using the second person, imperative language... Sure. Fully grown tech savvy people expecting it to follow detailed and nuanced instructions? That's pure comedy fodder I guess.
sltkr 3y ago
It's jarring because you can give it an instruction like "I want you to talk like a pirate from now on" or "Please end every reply with 'beep boop'" and it will happily oblige and continue to do what was asked continually. So it's strange that being apologetic is apparently so ingrained in its training that it cannot stop apologizing even when you specifically requested that it refrained from doing so.
- namaria 3y ago
  I think it's more strange that the coincidence of some imperative prompts generating texts that looks like it's answering to it makes intelligent people think it's reasoning about the instructions and responding meaningfully to them instead of, you know, generating text as it was programmed to do.
- HarHarVeryFunny 3y ago
  It treats the system prompt specially - it's more than just a prefix to the conversation. OpenAI have said they put a lot of work it making it follow it as closely as possible (but no details of how they did it).

18 ms·

I don't understand why people insist in giving instructions to a chatbot when we know it works by generating texts based on prompts.

by namaria 3y ago

wizofaus 3y ago
You might as well ask why people bother asking it questions or talking to it at all on that basis. The fact it does typically respond differently to commands than to questions is part of the power of how LLMs work (whether that's entirely due to the fact its training data includes examples of commands being given and appropriate responses or it's explicitly part of the programming I couldn't venture to say, though I suspect the former).
- namaria 3y ago
  Talking to it... I get. It is a chat bot. Using the second person, imperative language... Sure. Fully grown tech savvy people expecting it to follow detailed and nuanced instructions? That's pure comedy fodder I guess.
sltkr 3y ago
It's jarring because you can give it an instruction like "I want you to talk like a pirate from now on" or "Please end every reply with 'beep boop'" and it will happily oblige and continue to do what was asked continually. So it's strange that being apologetic is apparently so ingrained in its training that it cannot stop apologizing even when you specifically requested that it refrained from doing so.
- namaria 3y ago
  I think it's more strange that the coincidence of some imperative prompts generating texts that looks like it's answering to it makes intelligent people think it's reasoning about the instructions and responding meaningfully to them instead of, you know, generating text as it was programmed to do.
- HarHarVeryFunny 3y ago
  It treats the system prompt specially - it's more than just a prefix to the conversation. OpenAI have said they put a lot of work it making it follow it as closely as possible (but no details of how they did it).