Verizon Launches AI Connect As McKinsey Predicts That 60-70% Of AI Workload Will Shift From Training To Inferencing By 2030

As McKinsey highlights a significant shift in how artificial intelligence will be used in the future, Verizon Communications Inc. VZ launched AI Connect during its fourth quarter conference call, to meet the growing demand for network capacity and computational power.

What Happened: Traditionally, AI models primarily focused on training, where they learn from massive datasets. Inference is the stage where the trained model is used to make predictions or decisions on new data.

According to a press release, McKinsey predicts that a substantial portion of about 60% to 70% of AI workloads will transition to real-time inference by 2030. This means AI will be used for immediate, on-the-spot tasks rather than just offline analysis.

Verizon aims to become a key enabler of the AI ecosystem by providing the essential connectivity and infrastructure that AI applications rely upon.

McKinsey also outlines that there is an urgent need for low-latency connectivity, computing, and security at the edge beyond current demand.

• Low-Latency Connectivity: Real-time inference demands extremely fast data transfer. Edge computing, where processing happens closer to the data source (like a factory floor or a self-driving car), is crucial to minimize delays.
• Compute Power: Processing complex AI models requires significant computing power. This necessitates deploying powerful processing units directly at the edge locations.
• Enhanced Security: As more critical operations rely on AI at the edge, robust security measures are paramount to protect sensitive data and prevent disruptions.

In essence, the shift towards real-time inference from training will drive a surge in demand for advanced edge computing infrastructure. Companies will need to invest in high-speed networks, powerful edge devices, and robust security solutions to support this evolving AI landscape.

See Also: Trump’s Executive Orders Fuel Tech, Weaken Energy Stocks In His First Week

Why It Matters: Verizon, which reported earnings of $1.10 per share on revenue of $35.7 billion in Q4, met analysts’ expectations.

The telecom company has already funneled over $1 billion to leverage its existing infrastructure, said its business group CEO, Kyle Malady, during the earnings call.

Google, an Alphabet Inc. GOOG GOOGL company, and Meta Platforms Inc. META have purchased capacity in Verizon’s network with the intent of using it for their AI workloads. Players like Nvidia Corp. NVDA are also working with Verizon to reimagine how telecom functions can work along with AI workloads.

“The bottom line is we are ready to help power the AI ecosystem. We have the assets, the expertise and the vision to deliver AI solutions at scale. By unlocking the potential of our existing assets, we can further improve the financial profile of Verizon business,” said Malady.

Read Next:

• Nvidia, Microsoft, Or Oracle: Which Stock Among These Stargate AI Project Beneficiaries Are The Least Expensive To Own?

Photo courtesy: Shutterstock