Great piece: this is an excellent example of why critical thinking is a skill that never goes out of style. And while I find the conversational tone of AI to be concerning, I have to admit the line "why I was so remarkably wrong" did make me laugh.
Thank you! Yeah, it's very uncanny, someone else compared it to a conman pitch. It really does have that quality, but it was cool that it was also able to explain how it works (if that is indeed accurate).
This is a fascinating and insightful experiment. Thank you so much for creating and sharing it. FWIW, I always remind my ChatGPT "assistant" that I would like pithy responses. Not only are they easier to read, they also help focus my healthy skepticism of the reply. Cheers.
I'd say the "chatting" and constant idioms are less annoying and more disconcerting. The whole "conversation" (if it can be called that) sounded like a confidence man's swindle. Is this kind of assistance really what the data centers popping up all over NoVA are supposed to be supporting?
I did like Gemini's admission that it can't "see." We're so used to epistemologically regarding the internet in a human-sense way that it's easy to forget that these chatbots can't and won't regard the world in the same manner we do.
I enjoyed the general tenor of this piece, and skimming through Gemini’s attempts to justify its confident assertions in an increasingly desperate manner.
One thing though is the piece was difficult to wade through and follow cleanly because of all the bold headings in different sized fonts — I wonder if this would read more easily if your questions/prompts were in italics, while the Gemini responses were blocked off more authoritatively (in a visual manner) from your questions. Maybe Substack makes that difficult, I don’t know.
Great piece: this is an excellent example of why critical thinking is a skill that never goes out of style. And while I find the conversational tone of AI to be concerning, I have to admit the line "why I was so remarkably wrong" did make me laugh.
Thank you! Yeah, it's very uncanny, someone else compared it to a conman pitch. It really does have that quality, but it was cool that it was also able to explain how it works (if that is indeed accurate).
This reads like something from a Douglas Adams book. Especially the incompetent sycophancy of the AI. Share and enjoy!
This is a fascinating and insightful experiment. Thank you so much for creating and sharing it. FWIW, I always remind my ChatGPT "assistant" that I would like pithy responses. Not only are they easier to read, they also help focus my healthy skepticism of the reply. Cheers.
Thank you!
I'd say the "chatting" and constant idioms are less annoying and more disconcerting. The whole "conversation" (if it can be called that) sounded like a confidence man's swindle. Is this kind of assistance really what the data centers popping up all over NoVA are supposed to be supporting?
I did like Gemini's admission that it can't "see." We're so used to epistemologically regarding the internet in a human-sense way that it's easy to forget that these chatbots can't and won't regard the world in the same manner we do.
Yeah. It does have a very con-man-pitch feel to it. (Of course, maybe it really does "see" and is smart enough not to let that on - ha ha.)
If you're interested in more content that's critical of the AI industry, visit https://pivot-to-ai.com/
I enjoyed the general tenor of this piece, and skimming through Gemini’s attempts to justify its confident assertions in an increasingly desperate manner.
One thing though is the piece was difficult to wade through and follow cleanly because of all the bold headings in different sized fonts — I wonder if this would read more easily if your questions/prompts were in italics, while the Gemini responses were blocked off more authoritatively (in a visual manner) from your questions. Maybe Substack makes that difficult, I don’t know.
Thank you! Yeah, it is kind of difficult with the formatting because Gemini uses all these different headers too, which is annoying!