Anonymous #1 2d

Have you been on X recently? Boatloads of conservatives “@grok”ing the simplest things

8

Anonymous OP 2d

Yeah because it holds the world’s knowledge. That’s like saying someone is stupid for using google

0

Anonymous OP 2d

Lmao you have to pay for it now but I’ve seen some disappointing stuff

1

Anonymous #1 2d

At least with Google though you had to think about how to form a query that would get you the results you wanted, sift through sources, and think critically about where the info was coming from. LLMs do a lot of that work for you and just spoonfeed you the answer. Plus if you saw something on Twitter about like “Joe Biden shot dead in Gaza” and have to Google whether or not it’s true, I’d still think you’re dumb

4

Anonymous OP 2d

Why would you thjnk the person is dumb. That’s something that can actually happen.

1

Anonymous #1 2d

Because it’s highly implausible and even in the wild chance that happened, I doubt you’d hear it first from like MAGA AMERICA X NEWS on social media

6

Anonymous #1 2d

Frankly another big issue of this is people falsely believing that AI has all the answers and is correct all the time when it demonstrably is not. It provides false responses all the time. One of the most obvious examples was Grok not believing Kirk was dead. Or when Grok’s output was intentionally modified to say Elon was the best person ever at any physical or mental activity.

1

Anonymous #1 2d

And frankly I also do think that people’s over-reliance on Google was already a problem before AI made it so much worse. People used the first search result rather than knowing how to do actual research, and disinformation was spread that way all the time.

1

Anonymous #3 2d

The Grok not knowing about Kirk dying was definitely because the model was trained before that happened. But the Elon dickriding was genuinely insane 😭 just shows that it’s all about the training data, the parameters, and the system prompt

1

Anonymous #2 2d

And it shows how insidious this reliance on AI is. People assume it’s objectively correct and holds all the answers. Then the company that owns the AI can just modify the output to push whatever agenda they want. Or the output is unintentionally biased just from the training data.

5