I was going to just leave a comment on your post, but I thought this was an intriguing enough idea that I would actually create a post withy my analysis.
Presentation Comments: I actaully was able to watch this demo yesterday, and it was absolutely hysterical. It was definitely entertaining. The demo was impressive, and the presenter was very charismatic and funny. There was quite a bit of language barrier there, so it’s hard to say how developed their concept truly is from the Q&A section.
Feasibility Comments: (Let me begin by first saying, keep in mind that there demo is not anywhere close to a fully functional prototype. As an example, for my senior capstone project, we created an X10-based home video alarm system (uses powerlines to communite signals through radio frequencies) that interfaced with desktops and PDAs via the internet and could be monitored mobily by the user or a 3rd party. It was far cheaper than commercial grade standard equipment. In concept it was a really cool idea that would have taken a bit of funding to work out developmental issues that we had with our crude prototype, but with limited resources we were able to create a “demo” that made it look really kick ass.) I really like the concept of the technology, but I remain skeptical that what was shown in the demo can become a usable application in the next several years for several reasons.
Sekai Camera feasibility comments:
- The key to making this work is object recognition AI, and that is still a little bit away as I have seen MIT demos of this type of technology, and it is very crude right now. Without it, like one of the commenters brought up, you would just have lists that generated based on positional data. This would require constant updating by users or a 3rd party.
- Even if this AI technolgy is developed right away, modern day cell phones don’t yet have the computational capacity to run object AI algorithms of this complexity yet. There are only very crude versions of it now, like recognizing a “face” on a body in facebook.
- Getting involved in the AI scene is way over my head. The field of AI is extremely interesting to me as I was able to interact with some people in college that were invovled in some projects, but they are so complex it’s not even funny.
- The area that will be essential and more doable is information agrregation and presentation. Similar to what social network and information aggregators are doing will have to be done to the extreme amount of geographic, product, business and internet data that would need to be access and preseneted on the fly by Sekai Camera.
IdeaFLITE 