8 ms·

Worse, how would a GPT be "biased" in the first place? To claim it is biased implies OpenAI did this on purpose, there is no generous way to read this that indi

Worse, how would a GPT be "biased" in the first place? To claim it is biased implies OpenAI did this on purpose, there is no generous way to read this that indicates UEA did this study in a neutral way. ChatGPT was trained on, essentially, everything on the Internet... if ChatGPT has a "bias", then, no, it doesn't: humanity does.

127 3y ago
Any "alignment" is adding bias and reducing usability of a LLM. You are trying to wrestle the model towards what you find appropriate instead what is the average of the source material.
- perfmode 3y ago
  The bias could be in the source material no? Or in the accidental choice of source material no?
redeeman 3y ago
no, openai had an army of people to classify things
wtallis 3y ago
> To claim it is biased implies OpenAI did this on purpose You're relying on a particularly unhelpful and narrow definition of "bias". There are tons of valid and widely-accepted uses of the word to apply ot unintentional bias and to apply to things incapable of having intentions.
- tstrimple 3y ago
  Very true. Facial recognition models which don't work very well on dark skin aren't that way because the creators of the model didn't want it to work on folks with dark skin. It's because of the bias built into the training material. Those models are biased.
archon1410 3y ago
> To claim it is biased implies OpenAI did this on purpose They did do it on the purpose—the RLHF. The RLHF process was definitely biased, deliberately or unconsciously, in favour of certain political, social and philosophical positions (apparently the same ones held by many in Silicon valley/OpenAI teams responsible for this).
- jackmott42 3y ago
  Good, we don't need the robots being ignorant, racist, insurrectionists also.
jackmott42 3y ago
Current conservative americans have already noted amongst themselves that their policies are problematic because reality has a liberal bias.
- BitwiseFool 3y ago
  >"because reality has a liberal bias." An expression which is only said by those with a liberal bias.

127 3y ago

Any "alignment" is adding bias and reducing usability of a LLM. You are trying to wrestle the model towards what you find appropriate instead what is the average of the source material.

perfmode 3y ago
The bias could be in the source material no? Or in the accidental choice of source material no?

redeeman 3y ago

no, openai had an army of people to classify things

wtallis 3y ago

> To claim it is biased implies OpenAI did this on purpose You're relying on a particularly unhelpful and narrow definition of "bias". There are tons of valid and widely-accepted uses of the word to apply ot unintentional bias and to apply to things incapable of having intentions.

tstrimple 3y ago
Very true. Facial recognition models which don't work very well on dark skin aren't that way because the creators of the model didn't want it to work on folks with dark skin. It's because of the bias built into the training material. Those models are biased.

archon1410 3y ago

> To claim it is biased implies OpenAI did this on purpose They did do it on the purpose—the RLHF. The RLHF process was definitely biased, deliberately or unconsciously, in favour of certain political, social and philosophical positions (apparently the same ones held by many in Silicon valley/OpenAI teams responsible for this).

jackmott42 3y ago
Good, we don't need the robots being ignorant, racist, insurrectionists also.

jackmott42 3y ago

Current conservative americans have already noted amongst themselves that their policies are problematic because reality has a liberal bias.

BitwiseFool 3y ago
>"because reality has a liberal bias." An expression which is only said by those with a liberal bias.