What happens when a veteran network architect can't get his own Wi-Fi working before a podcast? Exactly the kind of conversation you can't plan for.

Victor Kuarsingh and I have known each other for nearly 20 years, going back to when Jason was at Juniper and Victor was designing IP architecture at Rogers. Victor is now the Managing Vice President of Connectivity at Capital One, and Jason has a little podcast.

In this episode, we pick up where that era left off and look squarely at where things are actually heading.

The conversation moves fast: from managed Wi-Fi and proactive network maintenance, to AI agents running live infrastructure without human approval, to the deeply uncomfortable question of whether anyone is building these systems with a real off switch.

No pitches. No prepared talking points. Two people who have spent decades inside the infrastructure everyone else takes for granted, being candid about what is exciting, what is broken, and what keeps us up at night.

We discuss

Home networking reliability — the irony of a network guy struggling with his own Wi-Fi setup, and the broader point that reliability has become more important than raw speed

AI for network observability — using AI to monitor, diagnose, and self-heal network issues, and how much autonomy you give it before humans need to stay in the loop

AI guardrails and the "how much rope" problem — where autonomous AI decision-making is genuinely better than humans, and where judgment calls still require a person

Personal AI use and critical thinking — how you're both using LLMs day-to-day, the risk of people outsourcing thinking entirely, and why cross-referencing outputs matters

Financial services as the next big disruption — Victor's view that fintech/payments is where telecom was 20 years ago, ripe for the same kind of transformation

Digital identity and voice authentication — deepfakes, voice cloning, and whether we'll ever have a reliable way to confirm who we're actually talking to

The "just because you can, doesn't mean you should" question — AI existential risk framed through sci-fi (WarGames, Battlestar Galactica, Demon Seed), wrapping up with humanity's tendency to build first and ask questions later

Enjoy!

You can find Victor on LinkedIn

Make sure to check out the Jason's Industry Insight Newsletter