Is IT Ready for AI?
During a particularly engaging session on modernizing IT infrastructure for AI readiness at the edge, an impressive panel of experts gathered to deliberate on the topic, including: Mark Romanowski, EVP, general manager, Agilant Solutions; Mike Trojecki, senior director, AI Practice, WWT; Jon Gould, director of business development, software, Schneider Electric; and host Gordon Lord, VP, US Channels, Schneider Electric.

The consensus from the group was essentially that AI will reshape everything we’ve come to know, and we are only at the beginning. It’s on us now to ensure that we do so responsibly and ethically, and if we do this while also ensuring sufficient data collection, edge computing, and simulations techniques, the opportunities will be endless.

“AI is a journey. It’s not something that is going to be quick, but it’s extremely transformative. What we are seeing now is nothing; what we are building toward is purely Star Trek,” said Romanowski. “It [AI] is magnitudes bigger than the internet or digitalization. [The] one word answer is ‘opportunity’ in terms of what’s most exciting in AI today,” said Romanowski.

The tools behind what make AI were also a discussion point, particularly in terms of how they can help streamline AI application developments, how customers can apply AI in a lightweight way to drive value and innovations in their own operations, and how partners can help facilitate this.

“In the next two years, AI at the edge will grow more significantly,” said Gould. “Inference will go out more toward the edge, and you’ll find that out in the channel, the scope will grow to facilitate AI expansion, primarily on the enterprise side.”

Processing Power and Simulation
Advances in AI are only possible through innovations happening elsewhere, such as in industrial computing and processing power, which in turn help with AI simulation and training. One term the panel called out is quantum processing units (QPUs), which could be one thousand times more powerful than GPUs, suggested Romanowski.

“Companies will have choices, which could include CPUs, GPUs, Google’s TPUs, and eventually, QPUs, when it comes to the edge,” said Trojecki.

Such advances will significantly benefit simulation and digital twin technologies, such as NVIDIA’s Omniverse ecosystem, which has helped AI technology take huge leaps forward. In a fireside chat between Pankaj Sharma, EVP, Secure Power & Data Centers Division and Global Services Business, Schneider Electric and Rev Lebaredian, VP, Omniverse and Simulation Technology, NVIDIA, Omniverse was discussed at length. During the talk, Lebaredian called Omniverse a “renewable physical data generator” where AI can be tested and validated. Before deploying, companies want to make sure there isn’t a bug or a regression in its behavior, so they can use simulation to do extensive testing using the libraries, SDKs, APIs, and specialized generative physical AI algorithms NVIDIA has compiled within the Omniverse platform.

“We can generate infinite amounts of data about the physical world, and these simulations can be the training grounds for AI that will operate in the physical world,” he said.

Expect a Surge in Physical AI
The topic of physical AI, which NVIDIA defines as the ability to let autonomous systems such as robots, cameras, or vehicles perceive, understand, reason, and perform or orchestrate complex actions in the physical world, was discussed in several places, including the panel.

“Physical AI is the next wave of edge applications. As a company we’re seeing many requests for integrating AI into robotics. You don’t have to have a massive infrastructure for these applications,” said Trojecki.

Advances are also being made when it comes to the intersection of physical and generative AI, as well as simulation, according to Lebaredian.

“We are now in the AI phase where we’re talking generative AI models, giving them physical embodiments, whether it’s a robot or a vehicle, and giving them a way to sense the environment and physical world through sensors like cameras or lidar, allowing them to make decisions and act inside the real world,” he explained. “Gathering data from the real world in most cases ends up being too expensive, however, so we take the laws of the physical world, bring them into a computer, and use simulation to test and validate AI.

Only the Beginning
The idea that we are only getting started in terms of how AI is going to add value on the manufacturing floor and beyond is staggering, considering how prevalent it is becoming, but this was indeed the sentiment of many.

“What is most exciting in AI is that we haven’t even seen what’s coming yet,” said Trojecki. “It feels like my job changes every three weeks. AI will change everything about humanity, but we need to do it ethically and responsibly, and this is what is most exciting.”

Need Help Navigating AI?
Navigating the marketing landscape in the age of AI can be challenging. If you have questions about optimizing your website for generative AI, need useful generative content on AI, or are looking to elevate your marketing strategy, TECH B2B Marketing is here to help. Contact us today.

Navigating the marketing landscape in the age of AI can be challenging. We can help. Contact us today!