[Brian]

I think when you put all of his statements and hints together, the following is likely:

What are the implications of a world where everyone has a super high-speed Internet connection in their pocket and many gigabytes of storage, super-fast processors, audio, visual and multimedia?

1. The upcoming device will probably be a Treo/LifeDrive hybrid, probably with Wifi and Bluetooth, and 4-6GB plus of internal storage.

Q: What exactly are you building, then?

A: I'll try to make it as simple as possible. Ultimately, it's the algorithm of the neocortex, the human brain. It lets you build machines that do things that humans can do. We can see what things are. We can understand language and we can type on a computer and move about. This is all done using this memory system that's in the brain and we can now build this memory system in software or in hardware. Our products are a set of tools that allow you to configure these memory systems, which we call HTM, hierarchical temporal memory. You can interface them with a thing like a camera or a microphone or sonar and it learns about its environment in the same way you learned about your environment when you were a child. It can model the environment, recognize things in it and make predictions about the future.

What we're building is actually a platform. It's like a new type of operating system. It's a platform on which people can use our tools to create new applications for solving different types of problems.

Q: You could build it into a digital camera that would recognize a scene?

A: You could do that. You could have a computer that looks at images and knows what they are. It could look at images of Janet Rae-Dupree and know who you are. A car manufacturer may make a car that understands traffic. You could have a car that knows how to drive or that recognizes dangerous situations. You could have robots that can fix machinery. It's a memory system that learns the causes in the world. So what is a cause? Causes are things that actually exist. They're persistent objects like a car or a train or a desk or a building that present themselves to your brain through your senses. What the eye actually sees and the ear hears is really messy. The unsolved problem in artificial intelligence is: Given what's on your eye, what's really out there? You could call that object recognition but it's more sophisticated than that. Causes can be physical things, but they also can be things like ideas and words, music is a cause, a song is a cause -- anything that persists in the world. The brain doesn't know at first what's out there. When you're born, you don't know anything about trains and cars and buildings and language and so forth. The brain discovers all of that.

Q: How soon might we see HTM applied to real-world problems, like hurricane tracking?

A: Hard to say. I would say that it should be in five years but I wouldn't be surprised if it happened a lot sooner. I think it's more like two or three years.

2. Based on the timeline of his predictions for applying "real intelligence" to actual products combined with his statement in the Analyst Meeting that this new revolutionary device will "keep the company going over the next half decade", whatever this new device is, it will be the predecessor to a mobile computing device that has Numenta's "real intelligence" operating system.