AI: Help or hype for the photonics industry?

The inaugural Global Photonics Economic Forum, organized by Optica, took place at the beginning of the month and drew more than 300 leaders from the global photonics industry, including corporate executives, political figures, and investors, who engaged in meaningful conversations about the future of photonics business.

As we discussed its strategic direction, the topic of generative artificial intelligence (AI) and its impacts on our industry repeatedly rose to the surface. Whether during a CEO roundtable panel exploring the direction of the photonics industry, a keynote on photonics solutions enabling advanced packaging in the AI era, or general breakout discussions, we continuously came back to an examination of AI’s potential in our space.

Yet, as diverse as the dialogue was, much of our focus remained on debating whether we are falling victim to shiny new technology syndrome or if AI will lead to true transformation. Ultimately, this discussion drove two overarching questions: Is this level of focus sustainable? And how it will influence the direction of the photonics industry?

Behind AI’s potential

While these are layered questions, the data shows companies are willing to bet on AI’s staying power. Beyond data center investments, which drove analyst firm OMDIA to refer to AI demand as a “new industrial revolution,” reports predict more than $70B in revenue by 2027 for the planned compute capacity required to support the need for generative AI (see OMDIA’s Optical Communications (OFC2024 and AI) Report; available to Optica Corporate Members).

And new applications heightening that demand continue to emerge. For instance, in a recent interview at Stanford University, former Google CEO Eric Schmidt pointed to AI’s technological prowess. He highlighted context windows as a powerful tool with long-term memory capabilities, similar to human short-term memory. Schmidt also introduced the concept of large language models (LLMs) as agents that can read and understand complex topics. During the discussion, he explored a hypothetical scenario where TikTok was banned, and users created a similar platform in 30 seconds using LLM agents.

Bursting the bubble

But this level of intensity harkens back to the late '90s/early 2000s and a similar obsession with telecom—and we all know how that ended. So we need to at least consider the possibility of AI moving in a similar direction. Will AI be its own bubble?

Not according to leaders in the field who think we’re just scratching the surface of AI’s potential. For example, when I posed this question to Michael Lebby, CEO of Lightwave Logic, he said AI’s here to stay. And he pointed to the residual impacts AI’s growth will have on photonics.

Take OMDIA’s point, as expressed by Lebby: If AI generates more computational processing, it, in turn, will mean more use of memory locations on GPU-type chips. For these computational processes that need to be sent off chip, we may look to the fiber-optic network through data centers and telecom, and this means more information must be sent much faster at a lower power consumption. Ultimately, the demand for AI-based solutions will drive new technologies and photonic solutions to enable them. Lebby believes that, given this demand swell, it’s very likely optoelectronics will be a strong competitor to the copper interconnects being used today.

Taking AI into the future

But to unlock its full potential, AI will need a flexible regulatory environment to enable cautious exploration. In his talk, Schmidt emphasized that Europe’s stringent regulatory approach, particularly around data privacy and AI ethics, could either serve as a model for responsible AI deployment globally or hinder its competitiveness by stifling innovation. Schmidt specifically calls out the General Data Protection Regulation (GDPR) as a double-edged sword: While it sets high standards for data privacy, it also imposes significant compliance costs that can slow down AI advancements.

Because with any new technological evolution, there’s a balance between risk and opportunity. And with AI, ethical considerations add a level of complexity. No one wants to unleash Pandora’s box, so that leaves AI as a technology that will spark continued debate, investment, and research from the photonics industry and beyond.

As new potential emerges, Optica Corporate Members will be at the forefront of these advances, driving what’s next. And with AI’s apparent staying power, you can bet they will live up to the hype.