In 1998, 48% of respondents in one survey said they never used the internet. Just a few years later, weekly use was growing more normal. Now, it’s everywhere, all the time.
We already saw Grok being used for this, though rather clumsily by stuffing the prompt.
If an AI company were behind the scenes fine tuning on specific political sentiment you would never know.
In fact there’s some evidence that later ChatGPT models are more right wing biased than early models (which were accused of being left wing).
Also important to note how much social media gets fed into these things and how astroturfed modern social media is these days, so even if not explicitly biased the well has been poisoned.
We already saw Grok being used for this, though rather clumsily by stuffing the prompt.
If an AI company were behind the scenes fine tuning on specific political sentiment you would never know.
In fact there’s some evidence that later ChatGPT models are more right wing biased than early models (which were accused of being left wing).
Also important to note how much social media gets fed into these things and how astroturfed modern social media is these days, so even if not explicitly biased the well has been poisoned.