Now Reading:

Artificial intelligence and the chip industry

By Udita Shukla 

There is a lot of noise around the advent of autonomous cars in the near future. Industry experts have been harping on about how self-driving cars would be plying the roads in the not-so-distant future.

The market for chip tech

Intelligent transport systems demand powerful chips that can trigger responses in real-time and dynamically evolving environments. For instance, a fully autonomous car will have to rapidly update its position and crowd density around it with rapidly changing traffic conditions and pedestrians walking down the street. In order to fulfil one response to similar external stimuli, the process involves inputs from the body-mounted camera, radar and LIDAR (Light Detection and Ranging) systems to be assimilated into the sensors and subsequent identification of objects to ‘decide’ upon the next move of the vehicle. Therefore, the automotive market is one of the major industries that has created opportunities for disruptions in chip technologies to cater to the fast speed and response-times of current and future automobiles.

AI taking over the world

Artificial Intelligence (AI)  algorithms are able to teach themselves the complex tasks of decision-making as close to a human being as possible. Given this eerily exceptional ability, they form the ‘nervous system’ of chips being currently developed and worked on, to run the gadgets and intelligent transport systems, and, scientific and healthcare domains of tomorrow. Software giants, namely, Microsoft, Google, Baidu, etc. invest billions of dollars and intellectual property to train their deep neural sets. They intend to come up with the perfect chip which can catalyse both training and execution of miscellaneous technological systems.

Google’s CEO, Sundar Pichai, has been conspicuously vocal about the meagre concentration of people in the world who actually qualify for being able to build machine learning models. This was shortly followed by the introduction of AutoML—Automated Machine Learning model Google has been working on, that apparently transcends the boundaries of existing hand-built ones, and also consumes less processing power. Summarily, Google has more or less confirmed its perception of the future—artificial intelligence combined with the internet of things. This means our world will be shared by the likes of Pixels, Google Homes, Google Assistants, even beyond!

Google’s faith in AI

The differentiating feature in Pixel phones is the capability to detect objects competently enough to reproduce the same quality while employing a single camera. This is a great edge over Apple phones which have two cameras to produce the same effect in photographs. This feat is the gift of A.I. that works (quite literally) behind the scenes.

One of the most disruptive offerings by Google is its alternative to Apple AirPods, namely, Pixel Buds which not only boast the complete functionality of Bluetooth earphones but also facilitate almost real-time translation via Google Translate. So much for destroying language barriers and connecting people! Although this feature will only work when coupled with a Pixel phone, the Assistant (on the Pixel phone) can also access Google Lens, thereby, facilitating search results depending on what is ‘shown’ to the camera.

Pervasiveness of AI

As astounding as these technological enhancements are, something that cannot escape our eye is the indispensability of A.I., and its silent integration in our daily ecosystem. Artificial intelligence has managed to trickle down the most restricted crevices of the socio-economic, political and societal areas of the world order. With ever-powerful chip and processing technologies, the data manipulation and analysis speeds have defied all limits of conventional computing and sensing.

The possibility a world driven by data and its optimisation has taken ginormous proportions in terms of ever-maturing deep learning and neural network technologies. Datasets are growing at the rate of exabytes per year, while, machine learning algorithms and neural networks are being ‘tamed’ to identify, remember and co-relate to our real world as closely as ever.

Sure enough, the future is ‘artificial’ and ‘intelligent’, with a tiny chip at the heart of everything.

Featured Image Source: Pexels

Leave a Reply

Your email address will not be published. Required fields are marked *

Input your search keywords and press Enter.