AI everywhere: artificial intelligence, neural interfaces and holodecks
Pervasiveness is a difficult thing to achieve for any technologies, that too for a sustained time period. However, when something achieves that, it usually goes on to become the zeitgeist of a time. AI or artificial intelligence is such a technology, as the current trend seems to convey. Nvidia, a company that was once a household name for its fantastic graphics cards at affordable price, has been at the centre of the AI whirlpool for some time now. Currently, surprising as it may seem, its AI involvement goes way beyond its initial concern and the most recent investment has been the hot field of automated driving.
Now, as they proclaim it to be, AI has the power to steer future in a particular direction altogether and Nvidia is taking on graphics designing in a new mode altogether by creating futuristic GPUs. Since Nvidia has been investing in GPUs from the very beginning, their strides to make GPU smarter, faster and more efficient has laid the foundation for AI to arrive and it is why Nvidia is reaping benefits currently. More importantly, Nvidia wants to spread this effort across the globe and hence, they have started becoming angel investors of many projects around the globe. However, the change has been in the recent decade.
The change in the last decade
In the recent years, the obsession was with the basics as Nvidia took one step at a time. They focused on simulation of physical quantities and graphics qualities during this time. Some of these concerns were fluid simulations, finite element analysis or even molecular dynamics. Basically, they were working on Newtonian physics and its many aspects in daily lives. However, things changed quickly once the basics were taken care of and this change has become the trailblazer of Nvidia and its current success.
Moore’s law, as IT companies continuously proclaim, is no longer producing technology at the rate promised. While they have devised accelerated computing that have benefited the evolution until now, it is no longer moving as fast as Moore’s law claimed it to be. Now that computer graphics, augmented reality and virtual reality have come together to create a heady mix of ideas, simulations have also evolved as new and extremely challenging simulations of physical quantities are being stirred up now and then.
Nvidia, obvious as it may seem, are now operating on supercomputers, trying to harness technology that were once out of reach, and AI is one of them. The arrival of deep learning in technology totally shifted the scene and Nvidia took quickly to it. Combined with the current processing skills powered by Nvidia, deep learning could be recognized as the watershed moment in AI.
Why has it been at the centre?
Despite some alternatives, GPU still rule the AI universe and it is precisely why GTC, the annual conference of Nvidia attracts millions of researchers around the globe. In fact, Nvidia has created a program named inception so that it can support start-ups gearing up for AI. In short, it wants to expose the IT world to AI more and more in form of investment, learning, expert advice and many more. People from completely alien industries are coming in these events to understand how AI can impact lives beyond simple computer science and become a household phenomenon.
Nvidia believes that AI will replace software as it does the same thing with much more efficiency and intelligence. It can perceive and plan according to the given scenario and needs no intervention to change its course. In fact, software’s possible direction will also be dictated by AI in the future as machine learning is continuously adopted in this framework. It will also not be limited to cloud-based AI services and data centers as proclaimed.
There are so many hardware objects from cars to microphones to drones and so many devices to integrate deep learning, that AI offers infinite sensors that will collect enormous amount of data. Nvidia imagines a future where every device that can transmit some sort of data will have an interface for deep learning, resulting in trillions of devices. In fact, gaming and visualization are also part of the plan and they are trying to incorporate some of it to make gaming interactive to the extreme.