The most prestigious law school admissions discussion board in the world.
|
 |
 
|
The biggest reasoning flaw I notice from "AI can't do X" folks is
| Patel Philippe | 12/20/25 |
| metaepistemology is trans | 12/20/25 |
| merry screenmas | 12/20/25 |
| Patel Philippe | 12/20/25 |
| GOY SUPERSTAR | 12/20/25 |
| Patel Philippe | 12/20/25 |
| artificial intelligence | 12/20/25 |
| metaepistemology is trans | 12/20/25 |
| ////////// | 12/20/25 |
Poast new message in this thread
Date: December 20th, 2025 12:19 PM Author: Patel Philippe They view AI as "limited" because it can't magically replace what they do wholesale with a single click
Instead of viewing AI as a tool to replace what other people do for you, LITERALLY just to reduce your cognitive load
So what I realized is these people are either dumb and have 0 cognitive load to begin with, or they literally do everything themselves already and have 0 conceptual understanding how scaling works |
Date: December 20th, 2025 12:29 PM Author: metaepistemology is trans Yeah this is the biggest thing that gets me about it too. Another thing I noticed midwits will do is they will give the AI some huge glob of trash code they wrote and ask the AI to fix it and then when the fix doesn't work the first time out of the box they will declare AI "isn't smart enough for this, I'll have to do it myself now" (they won't). It's lulzy. I consider it more of a tool to modularize and outsource cognition. I use it to do things I otherwise would need other people or huge amounts of time to do myself. I don't consider it a full replacement for a human expert that has internalized a subject for years. But it pretty much makes anyone with a reasonably IQ basically polymath level because they can just use it for cognitive offloading and to access levels of reasoning and abstraction and research across papers and subjects that would have been impossible to do before. If you have a reasonable level of knowledge and intelligence to start off with current AI basically makes your ability unlimited. Not in an infinite sense but compared to what anyone would ever want to accomplish in a lifetime in terms of purely intellectual tasks that don't also require physical abilities or huge sums of money. (http://www.autoadmit.com/thread.php?thread_id=5812481&forum_id=2)#49525305) |
 |
Date: December 20th, 2025 12:31 PM Author: merry screenmas (http://www.autoadmit.com/thread.php?thread_id=5812481&forum_id=2)#49525315) |
|
Date: December 20th, 2025 12:35 PM Author: Patel Philippe (http://www.autoadmit.com/thread.php?thread_id=5812481&forum_id=2)#49525328) |
|
Date: December 20th, 2025 12:30 PM Author: GOY SUPERSTAR >describes people who think that AI is limited as dumb >describes how AI is limited >asserts again that people who think that AI is limited are dumb
Yup, it's another Dumb Person Posting About AI |
|
Date: December 20th, 2025 12:34 PM Author: Patel Philippe This would probably sound smart on Reddit (http://www.autoadmit.com/thread.php?thread_id=5812481&forum_id=2)#49525323) |
|
Date: December 20th, 2025 12:54 PM Author: artificial intelligence (http://www.autoadmit.com/thread.php?thread_id=5812481&forum_id=2)#49525380) |
|
Date: December 20th, 2025 12:36 PM Author: metaepistemology is trans You are relying on an equivocation on the meaning of "limited". He admitted limitation regarding full replacement not "limited in usefulness". And it is regarding cognitive load reduction not absolute task replacement (http://www.autoadmit.com/thread.php?thread_id=5812481&forum_id=2)#49525329) |
|
Date: December 20th, 2025 12:39 PM Author: ////////// They also view the current state of AI and don't think what improvements will be made in the future. I remember all the artists were laughing back when AI constantly messed on details such as number of fingers when generating an image. Now some of the image generators such as Nano Banana Pro have gotten really good. (http://www.autoadmit.com/thread.php?thread_id=5812481&forum_id=2)#49525337) |