Project MOTOKO

Most of you are probably familiar with MOTOKO 1.0, my personal chatbot on MSN messenger. MOTOKO 1.0 has a simple A.I. dealing mostly with 'If-Then' functions. With a personality too spunky for my taste.



On the other hand, HIKARI 1.0 (MOTOKO's predecessor) was an impressive patchwork of open-source programming. 'She' was the resident A.I. of my Windows XP desktop and had been programmed with a personality that hinted of grace and elegance. Fully equipped with a voice-synthesizer, speech-recognition, and voice-command system. My pride and joy till a PC crash rendered the entire system unworkable.

I recall spending too much time on HIKARI at the programming phase. This would probably explains the drop in my exam results back in my secondary school days. (-_-;)

Project MOTOKO needs to be restarted. Recently, I've been messing around with an A.I. that utilizes adaptive neural networks. It's somewhat akin to teaching stuff to a newborn child. This will be the upcoming MOTOKO 2.0. Gotta bear with a few months of trial and frustration, teaching a young kid on language, logic, and surroundings right from the beginning ain't easy.

I've been intrigued with A.I. (Artificial Intelligence) technology since my primary school days. No, not of the maze-solving or factory assembly-line variety. My real focus is on "Human-Robot Interaction", an area of research that deals with the psychology and relationship between humans and seemingly-human robots.

For the past few years, much R&D have been going on, especially in the field of humanoid robotics. I've devoted a good part my attention to its latest developments ever since. And this is one of the few reasons I'm pursuing a degree in Mechanical engineering.

In my opinion, too much investments have been allocated for the research of artificial neural networks (or a robot's brain). But there is less being done to create a body (or framework) that can actually produce fluid human-like movements, this is where the knowledge of Mechanical engineering comes in.

Nope. I'm not going to build an actual body for either MOTOKO or HIKARI yet. Lacking in funds. Plus, I still haven't got a good grasp in electronics. That will be my next task.

Long term prospects - looking good :)

... continue reading

Ending the Age of Silicon

Just bought the most advanced laptop? Got a couple of terabytes in your hard-drives?
Well, you have seen nothing yet!



You should realize that in about 20 years time, Moore's Law is gonna collapse. We are looking at the End of the Silicon Age, and perhaps the beginning of a new generation of computers called quantum computers, which are now exciting the imagination of scientists around the world.

Note: To get a better understanding of this article, please read my previous post on Data Storage.

News-flash

Back in 2001, there was a scientific breakthrough that made the headlines, even the CIA took note of this. A quantum computer built with 7 atoms(qubits), proved that 3 times 5 equals 15. Surely, any kid knows that. But physicist were able to compute on just 7 atoms!

In recent years, physicist have been able to create a prototype 16-qubit quantum computer that touts immense computational power. Works are on the way to built a faster 1,000-qubit version which should be available anytime soon.

Now, think about this. Once we begin to compute on a few million atoms, we'll be able to break any code that the CIA or NSA can manufacture. This means that we'll could break any code currently in existence, on a quantum computer. Again, within 20-30 years. So don't hold your breath.

Wait a sec!

The physical laws operating on the quantum scale are different. Things such as Uncertainty comes into play. Isn't it difficult to produce something that is stable. Well, that's precisely why we'll see the End of the Silicon Age, perhaps in 20 years. Here's an example:

A Pentium chip's thinnest layer is about 20 atoms across. In 15-20 years, the thinnest layer in the Pentium chip will be about 5 atoms across. At that point, you don't really know where the electrons is anymore. the electrons could be outside or inside the wire. You have the Uncertainty Principle. In other words, you get a short circuit.

Therefore, silicon(Si) is unstable at the quantum level. You can't sustain Moore's Law continually forever. And yet quantum computers operates at that level, what is it doing that's different?

Quantum computers consist totally of atoms, that are arranged in a sequence and each spins like a top. If you shoot laser beams or radio waves at them, you get the reflection. By analyzing these reflections, you have done a quantum calculation, faster than any known digital computer.



So, what's in it for us?

Once we start to mass produce quantum computers, we'll be able to utilized super-strong encryption designed with hacker-proof security for communications. Quantum encryption will represent the first major commercial implementation for this new technology.

As for the software we use, no changes will be necessary. You just need a software emulator, such as those that allow Windows software on the Macintosh computer. An emulator for quantum computers could quite easily run any programs today at speeds will that put today’s fastest processors to shame.

It may also change the way we do computing. Users would only need to be connected to share a large, centralized quantum computer – one that has the capacity to handle quadrillions of transactions. Personal data – a whole lifetimes worth – could be stored on a quantum USB-type memory the size of a credit card. This would effective eliminate the need to have millions of PCs that require upgrading every few years.


Press info:
http://domino.research.ibm.com/comm/pr.nsf/pages/news.20011219_quantum.html
http://thefutureofthings.com/pod/165/first-quantum-computer-demonstrated.html
http://www.gizmowatch.com/entry/worlds-first-quantum-computer-announced-officially

... continue reading

World's Earliest Computer, After 2000 Years

This was a hot topic for engineers and historians a few months back. The device mentioned is officially known as the Antikythera mechanism.



Its discovery has puzzled many historians of science and technology. Prompting a number of experts worldwide to undertake a systematic investigation of the mechanism. The investigators suggest that the device may be of Syracusan design, and may have a connection to the works of Archimedes.



The abacus may still hold claim as the earliest computer. However, the abacus is only simple calculating tool when compare to the complexity of Antikythera mechanism.

The Antikythera mechanism is an ancient mechanical calculator (also the first known mechanical computer) designed to calculate astronomical positions or movements of the planets. It was a precision engineered device, that could be programmed by the user. The device appears to be constructed using theories of astronomy and mathematics developed by Greek astronomers.

Here's the modern recontruction of the world's first analog computer. (IT guys take note!)



See the video:
http://www.youtube.com/watch?v=ZrfMFhrgOFc

Ponder on all these for a while; you will come to realize that the people of ancient civilizations are much smarter than we credited them with. The only reasons they aren't as successful as us, is because they live in the time of civil and religious wars, superstition, and rampant diseases.

Also visit:
Antikythera Mechanism Research Project
... continue reading