So the takeaway is Howard = Harvard? Am I missing something or is it not unreasonable to think that different universities on average graduate candidates with different interests and of different calibers? sTeReOtYpEs and bIaS are just averages of real world data that people don't like. Also the OP asked ChatGPT to respond to an open ended question with no productive guidance and is throwing a fit when some of it isn't on target.
Why block out the name but leave the byline of “CEO of Textio”?That doesn’t exactly anonymize it. Lol.
Awesome job at complying with Blinds stupid policies but also revealing the facts!
She literally founded a company to push her DEI agenda.
now this is growth hacking!
Anyone who thinks bias is coded in a program is not someone you want to pay attention to!
If the input data is biased then the result is biaised. And ChatGPT input data is the internet, which is so biased. In this case the AI is not the problem, people are.
Even for the case of "bias" though, that implies that the data is wrong. It's not clear that's the case. The OP gave the AI a shit prompt and asked it to invent text without context. OP is literally asking it to hallucinate, so its output isn't evidence of bad data or even a bad model. Furthermore, again, the idea that graduates of Howard and Harvard are the same is bonkers. And that reality is reflected appropriately in both the data and the model. If you're asking the AI to hallucinate, I'd say it didn't make up an outlandish lie.
Or maybe reality is biased?
ChatGPT is just an "averaged union" of the knowledge on the Internet.
Which is quite dumb and biased
It is not chatgpt’s job to unbias the answer. It merely aggregates the knowledge on the internet which may be incorrect or biased. Feel free to do your own verification!
These people share cherry picked examples and then put a post as if they made some discovery lol