What Siri SHOULD be

The issue with things like that and evi is the integration, only siri is really easily accessible, everything else is restricted to apps.

 

What Siri SHOULD be?

It should be Siri. And it should be developing.

Just because another service has an extra feature set, Siri should simply be Nina?

It is well understood and Apple has seen enough from SiriProxy that Siri needs something like that. We all know that there is a Siri-App-SDK in the works. What should be surprising is that Apple did not release it with iOS 6. Well, lets wait for iOS 7 then.

 

I did watch the video? and i agree with you, the issue is that I think apple think siri is great as it is, and when then do finally add any new major features they will get all the credit when there have been others that did it long before.

----------

yay, yet another year...

 

Sometimes, I feel like I'm talking to kids who have basically no idea about about Artificial Intelligence or Machine Learning or Natural Language Processing. I don't want to brag about my PhD from CMU in Machine Learning or having worked at Stanford on Autonomous cars (well, I should stop), but it gets a bit kiddish when people simply want software to be developed by their standards. If you are thinking so seriously, just go out there and develop a 100% perfect piece of future-proof software that's complete by the standards of 2020.

It's hilarious how some of you young ones take software design, development, verification, manufacture-for-release for granted and think its a child play. Very basic concepts of software development for e.g. Agile Software Development are not even known and you people comment as if you know so much.

I may have been rude or highly critical but its disturbing to see kiddish people as yours claiming what a particular piece of software should have been.

 

I wish they'd figure out a way for Siri to detect an accent in English, like country or southern accent. I live in Florida, it takes me 2 or 3 times to have her understand me. It sucks!

 

By all means brag away. The problem I'm seeing is that the current voice technology is in the toddler phase. Everyone is trotting out "dog n pony" shows and the demos and the natives go "Wow". But what are we really getting excited about? A vast majority of the stuff people are demoing can be called from a simple search on a web page. The queries i'm seeing are too simplistic to be considered AI IMO.

I can ask my son (10) to do a series of tasks and he'll process and get the stuff done. At this point I cannot get Siri or any other service to handle more than a series of two tasks without crapping out.

Nina is no different here. Great demo in pulling data but still not seeing delegation and the ability to change contexts or provide options in the software.

What's needed is for someone to develop a hierarchy of tests in ascending complexity that can be tested on services so strengths and weaknesses can be known.

Right now untrained bloggers are focused on the minutiae of the technology (i.e which query completes a fraction of a second faster than the competition)

Here's hoping for big change.