There has been a great deal of hype as well as skepticism and confusion surrounding the metaverse concept.
For many, it has added to the confusion of an already elusive world of AR (augmented reality) and MR (mixed reality). But for the well-initiated, the birth of the metaverse marks a landmark moment in the extended digital world approaching the ‘second life’ that some people have long predicted.
News that the world’s top tech firms are developing AI supercomputers has further fueled this anticipation.
But what does the entry of supercomputers mean for the metaverse and virtual reality, and how can we manage it responsibly?
What Is A Supercomputer?
A supercomputer is a computer capable of high-level performance. That performance far outclasses any consumer laptop or desktop available on the shelves and can be used to process vast quantities of data, drawing key insights from it. These supercomputers are massive parallel arrangements of computers, or rather, processing units that can perform the most complex of computing operations.
When you hear about supercomputers, you’ll likely hear the term FLOPS, which stands for floating point operations per second. It is a key measure of performance for the said top-end processors.
Floating numbers are essentially those with decimal points, including the very long ones. They are key when processing large quantities of data or carrying out complex operations on a computer. This is where FLOPS comes in useful as a measurement tool; it tells us how well a computer will perform when managing these complicated calculations.
The Supercomputer Market
The supercomputer market is expected to increase at a compound annual growth rate of roughly 9.5% until 2026. Growing adoption of cloud computing, as well as cloud technologies, will fuel the growth, as will the need for systems capable of investing larger datasets in training and operating AI.
This industry has been booming in recent years. Landmark achievements have helped build public interest, and companies all over the world are striving to outcompete and outpace the competition with their own supercomputer projects.
In 2008, IBM’s Roadrunner became the first to break the one petaflop barrier. The system could process one quadrillion operations per second.
The Fugaku supercomputer, which is based in the RIKEN Centre for Computational Science in Kobe, Japan, holds a claim to being the world’s fastest machine, capable of processing 442 petaflops per second.
In late January, Meta announced that it would be developing an AI supercomputer, claiming it will one day be the world’s fastest supercomputer.
And what is its sole purpose? The project will enable the next generation of AI algorithms, the first phase of which is already complete. The second phase is expected to be finished by the end of 2022, at which point Meta’s supercomputer will contain 16,000 total GPUs. The company has said that it will be able to train AI systems with over one trillion parameters on data sets of up to one exabyte (or one thousand petabytes).
These numbers are impressive. But what does it mean for the future of AI?
Supercomputers Transforming The Metaverse: The World Ahead
The revolutionary uses of supercomputers include ultrafast gaming and instant and seamless translation of mind-bending quantities of text, images, and videos at once. It could be used to scan huge quantities of images or videos for harmful content, for instance, or identify an individual face within a huge crowd of people.
These computers will also be key in developing next-generation AI models. They will power the metaverse, laying the foundation upon which future metaverse technologies can rely.
But the implications of this power means there are serious ethical considerations for the use of supercomputers.
Ethics And AI
New technologies always demand societal conversations on how they should be used and how they should not, and supercomputers are no different in this regard.
AI has solved some large and complex problems in the world, but some flaws still remain. These are not caused by AI algorithms. Instead, they are a direct result of data that is fed into the AI systems. If the data has a bias, then the result of an AI calculation will carry that bias. If the metaverse and virtual reality really become a ‘second life’, are we bound to carry with us the flaws, prejudices, and biases that we have in the first life?
The age of AI brings with it key questions about human privacy and also the privacy of our thoughts.
To address this, we must seriously examine our interaction with AI by looking at its ethical structures to ensure that usage is non-bias, transparent, explainable, and accountable.
We must be able to explain why certain calculations or processes were initiated, what exactly happened when the AI ran it, that no initial human bias was against any group or idea, and who is to be held accountable for the results of these calculations.
It remains to be seen whether the supercomputers and the companies producing them will ensure that these key areas will be addressed consistently and transparently. But it will become more pressing as supercomputers wield more power and influence over our lives, both online and in the real world.
The surge from the supercomputing era will push the era of parallel computing and its use cases at the speed of thought. We see a future where supercomputers, together with intelligent software, will run on a hybrid cloud, feeding partial workflows of computation into a quantum computer – a form of computing that experts believe has the capacity to exceed even the fastest of supercomputers.
It remains to be seen how this coming era fuels the next generation of metaverse experiences as we move forward, but it certainly is exciting.