You can add location information to your Tweets, such as your city or precise location, from the web and via third-party applications. You always have the option to delete your Tweet location history. Learn more
Schematically, intelligence is skill divided by experience (I = S/E). Deep learning enables arbitrarily high skill levels, but requires insanely high amounts of "experience" (data) to achieve these levels, resulting in an extremely low intelligence factor.
Again, a DL model requires a dense sampling of what it's doing. An intelligent agent (like a human) can do extreme generalization from little data. At this time, no one has any clue how that works. However, it may not necessarily be very complicated. Who knows...
Who do you think is attacking the generalization problem correctly @fchollet ? Maybe @vicariousai ?
I know anyone using datasets does not.
The problem is focusing on the individual mind (whether human or AI). Humans not only know stuff, but we know who knows other stuff we don't know and we can ask. Our collective knowledge far exceeds our subjective knowledge; human intelligence is collaborative and social.
Human capabilities do come from our civilization -- externalized intelligence -- but even an individual human, although it is just a clever ape, is still many orders of magnitude more intelligent than AI today.
Yup, and several hundred million of years of evolution has no doubt helped. Animals such as deer - who can stand and walk almost immediately after birth - are a clear example of this. Our firmware is just really well trained to learn adapt to new situations.
Earlier researchers of AI, Douglas Hofstadter specifically, claimed analogy is the core of cognition. Our powerful generalisation ability is pattern-matching composite, abstractions, and tweaking one tiny detail to make it apply in the current situation. So *MORE* experience.
I haven't finished EGB yet but Hofstadter seems to be heavy on the loops theory. I think that optimism bias has a large role to play, humans get loads of stuff wrong but we just ignore it and keep on looping! Kahneman shows how fallible human "intuition" is in thinking fast & slo
Don’t remember who said “glorified nearest neighbor...”
@fchollet Mr. Francois Chollet I just wanted to know that how does a model embeds the context when we input text and image vectors like in case of say image caption generation.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
