I just published What does it mean for a machine to ‘understand’? https://link.medium.com/vvNZ4pks80 1/
-
-
Also, I'm curious what you would think of a trillion "if-then" statements that could respond correctly to a trillion questions. Is there understanding there in any meaningful sense?
-
I would say that it would understand only those trillion questions. If the Google folks could solve internet search with 1 trillion questions, I imagine they would be happy. But I'll bet there are always new questions. (And we want to do more than QA, e.g., dialogue, stories, &c)
- Još 3 druga odgovora
Novi razgovor -
-
-
Hold on. Understand is a word borrowed from humans. Why can't the ml community avoid using such words (or consent to being held to human standards)?
-
It is the word the critics use. But it is also our long-term goal. I advocate reducing "understanding' to functional tests that can be objectively measured. Lots of them.
- Još 3 druga odgovora
Novi razgovor -
-
-
Thanks for writing this, it's interesting. I think that you are strawmanning a bit though. e.g. I have never heard this argument seriously proposed: "Today’s deep learning-based systems don’t exhibit genuine understanding, and therefore deep learning should be abandoned"
-
Replace "deep learning should be abandoned" with "funding for deep learning should be cut", and you will hear it from many people in AI
- Još 2 druga odgovora
Novi razgovor -
-
-
Today's AI pretends to be intelligent, truth is people anthropomorphize its output. AI is all about building machines which can navigate purposefully, use and make tools and communicate meaningfully (call it Artificial General Intelligence if you like). We are not there yet.
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
The main criticism is the use of AI and 'intelligence' rather that more correct term ML. Hell, we don't have a good grasp of natural (aka human/animal) intelligence.
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.