SCTE seeks to make AI more useful for operators and technicians

  • SCTE will certify AI-powered data to help operators improve network operations and diagnostics
  • The trade group also plans to use AI to help technicians gain credentials while on the job
  • Thanks to AI, soft skills have become "the new hard skills," said SCTE CEO Maria Popo

As telcos hop onboard the artificial intelligence (AI) train, SCTE wants to ensure they’re not derailed by inaccurate data.

At SCTE TechExpo24, CEO Maria Popo announced SCTE has teamed up with CableLabs to provide AI certification for members, helping them improve AI troubleshooting and diagnostic agents.

Many operators are already using AI robot collaborators, “but do we trust these co-workers?” she said at a keynote. “We’ve all heard how off-the-shelf AI agents lie and hallucinate.”

SCTE will certify AI-powered data, derived from SCTE’s own standards and CableLabs specifications. This certification ensures that the AI systems trained using this data “meet industry-validated benchmarks,” Popo told Fierce.

Operators and vendors will eventually be able to integrate SCTE’s generative AI "embeddings" — representations of real-world objects — into their systems and tools “enhancing AI-augmented network operations and improving diagnostic accuracy.”

SCTE is collaborating with Comcast and other operators to figure out how to “bring the right data into the troubleshooting devices,” she said, so that when the technician is on-site, they can find the problem quickly and apply the right remedy. 

The goal is to train large language models (LLMs) to help operators improve diagnostic accuracy to almost 95%.

Not only is SCTE aiming to give operators “confidence that they’re training their organization on the right things,” she said, it’s using AI to ensure “there’s some kind of standard across what everybody’s learning.”

To that end, the trade group also wants to help broadband technicians improve their skills on the job. So, why not use AI for that too?

“Phase 2” of SCTE’s AI roadmap will involve embedding “certification logic within the AI data." This will allow technicians to accumulate “on-the-job micro-certifications” that automatically trigger based on real-world task completion, Popo explained.

The technician can then say, for example, they’re certified in fiber splicing after they’ve done it for a certain number of hours., without having to take a standardized exam.

“Why are we giving [technicians] standardized tests when there’s other ways for them to showcase what they know?” Popo said.

“What’s great about genAI [is] just by the questions they ask, it could probably tell what they need to learn,” she added. “We’re using the AI in a way that’s maybe more intuitive, where people learn naturally.”

Preparing workforce for the ‘experience era’

According to Popo, the “experience era” of broadband is poised to bring over 600,000 new jobs in the next seven years.

Experience in this case refers to operators going beyond speed metrics to focus on things like smart devices, telemetry and seamless connection across networks.

On the workforce development side, many operators are still doing hands-on training, she said. But a trend that’s emerging is how to “better integrate hard skills and technical soft skills together.”

"Because of AI, the soft skills are the new hard skills," Popo said. 

One potential soft skill is using AI for collaborative intelligence — not just for collaborating with the AI itself but for interacting with other humans. If an agent is speaking with a customer and can’t seem to reach a solution, the AI can make suggestions “if you ask it the right questions."

Popo also thinks workforce development will be geared more towards digital adaptability. Because there are so many digital tools available, people need to be able to “adapt easily from one tool to another.”

“People assume that high school students have a high aptitude because they’re always on social media, or they’ve grown up with phones and iPads." But that's not always the case, she said. “They know how to look through their phones but when they look at these tools, like where you collaborate or do task management, they’re lost.”