I started watching the video. By 2:30, I had seen a voice recognition Google search, and what appeared to be a voice recognition "I'm Feeling Lucky" Google search.<p>Bored with this, I started skipping around looking for the really whiz-bang AI. I saw some home automation stuff that didn't seem to be much more than something OS X Automator could do by kicking off a script.<p>If you roll enough dumb voice activated services together, that can still be a useful tool, but I didn't see anything that was at Siri's level. Maybe it's hidden in the video somewhere. If it is, it should have been showcased right away.
Hate to be negative, but not really all that impressed. JARVIS implies AI and conversational/contextual UI, this just appears to be a thin voice action skin which has been done for decades, not sure it was worthy of a Forbes article.
You have to click an icon to activate voice... I'm sure you can write a quick Python script to compare new audio input, stored in a watched directory, to a base audio dictionary that fires an API based on normalized matches to "terms" in that dictionary. If you say "search" the script compares/approximates to existing waveforms.<p>We should have a Speech-Computer Interface API, rather than boring ourselves with the standard "Search for...", that makes a canonical set of speech patterns which computers should respond to habitually.<p>Things like "I'm thinking..." (fires off search, feeling lucky, wikipedia, etc. mashup results pages), "I'm hungry..." (fires off search, geolocation information, personal biometric data, etc...), "Who is..." (wiki? pipl? facebook?), "What is..." (wiki?), "When is..." (shows a calendar)...
The Startup I am currently working on (also called "Jarvis Inc"), is aiming for the same star. While the current form is far from the contextual understanding. I love to see people working towards the same goal. This was our earlier mockup at jarvis.co. We only saw limited interest probably because of the high end market size. We are currently in the middle of building an open API platform as a spin off of this project.