![Miroyato [⧖]](https://pbs.twimg.com/profile_images/1153008781741187072/aUbvz1u2_normal.jpg)
Language understanding is ultimately about parsing sentences and paragraphs into intermediate representations & acting on them. Wolfram Alpha can do that (for some cases, not all see http://rebooting.AI ); GPT-2 can't. 2/2
Show this thread
GPT-2 v Wolfram Alpha Prediction is not understanding, and never will be. 1/2pic.twitter.com/KxC5xP6oac
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Intelligent agents optimizing for prediction without developing theories are purely associationist machines with no understanding. Ullman Tomer delves deeper into these issues from a computational and cognitive perspective. http://www.mit.edu/~tomeru/thesis/tomerThesis.pdf …
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
This is good. Tempting to make a Stephen Wolfram joke, but I’ll hold that back because genuinely this is good.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I think many have a specific & IMO implausible idea of what 'general' intelligence means, & so would say the former approach is not 'general' & never will be, while the latter is more properly 'general' so is more of a step towards AGI, even if it isn't very good yet.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
"Prediction is not understanding, and never will be." What does that even mean? If prediction is not understanding, it's 99% of it. The ability to predict events is what gives us a common sense understanding of our environment. We couldn't pour coffee into a cup without it.
- 1 more reply
New conversation -
-
-
It needs to be trained on real world video data, with floating labels over objects and their actions/interactions. In English.
- 1 more reply
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.