If you follow our blog, you know we have strong opinions about the impact that data-intensive AI workloads have on infrastructure which in many cases is already being pushed to the limit. Lately, more and more members of the larger IT community are beginning to sound the alarm about the sheer scale of the challenges involved, with much of the ensuing conversation focusing on data centers, and more specifically about not only how to exponentially increase the number of facilities online, but also how to power them.

But despite the incredible importance of data centers and the central role they play in IT infrastructure – see the article “Are We on the Verge of a Data Center Crisis,” by our CEO Charles Yeomans in Data Center Post – it is important to remember that they are not only crucially important pillars of today’s IT infrastructure impacted by the insatiable appetite of Large Language Models for more and more capacity.

The very chips and processors used within data centers, the “pipes” and cables that connect them, and the satellites that tie everything together – see Charles’ article “Rethinking Data Transmission to Unlock Efficient Satellite Communications” in Via Satellite – are also being pushed to the limit.

And just as importantly, AI is not the only transformative development in computing that requires enterprises to rethink how they move, use, store and secure data. As this article,  “The Unintended Consequences of the IoT’s Dramatic Growth,” in IoT Business News reveals, the sheer scale of connected devices is an important factor that will only grow in importance as machines grow more advanced and complex while simultaneously using AI.

All of this is to say that data-related needs, whether resulting from connected, increasingly smart devices or AI, are increasingly outgrowing the very infrastructure in all of its forms, wired and wireless, that makes it all possible. We have also entered a time when blind optimism and faith in our ability to innovate as we go will not suffice. Now is the time when only a new approach to data itself – one we are confident our Data-as-Codewords, Neurpac and our Persistent Cognitive Machine make possible – is needed more than ever. From this point forward, data itself must be a core building block in infrastructure as we know it.