New Zealand’s third-largest mobile operator 2degrees is ramping up its MVNO business in record time, using the cloud-based charging platform from Totogi (where I’m acting CEO). The operator selected Totogi's Charging-as-a-Service back in February 2024 to modernize its wholesale offering. Now, 18 months later, several MVNOs are already live on the platform. This is what a modern telco looks like! Any other vendor would still be sending invoices for “implementation phases.” With Totogi, 2degrees is already scaling up and capturing revenue. Woo hoo!
T-Mobile got hit with a $92 million fine for selling customer location data to third-party aggregators who resold it to bail bondsmen and private investigators. Here’s a tip, operators: if you’re looking for ways to monetize your location data, you don’t have to sell it to sketchy third parties and risk massive regulatory fines. Monetize it yourself using tools like Totogi’s PlanAI to deliver hyper-personalized offers to your subscribers—like targeted promotions based on real-time location patterns that actually drive revenue FOR YOU. Stop thinking like a data broker and start thinking like a modern telco who uses customer insights to boost ARPU.
Steve Saunders thinks telcos are smart to keep humans in the loop while Big Tech recklessly automates everything, but I think his take is totally wrong. Hyperscalers have spent the last 20 years automating the shit out of their data centers—because automating everything IS the only way to manage infrastructure at scale. Telcos say their own reluctance to automate is “strategic wisdom,” when it’s really just fear of change. There’s potentially an opportunity at the network edge for AI-driven applications, but that opportunity will go to the operators who can automate their networks, not those who think having humans manually manage network configurations is a competitive advantage. You can’t capture edge AI revenue when your own operations are still stuck in the Stone Age.
Google just signed deals for 2.3 gigawatts (GW) of nuclear power to handle AI data center demands that grew 17% last year alone. Remember when Roy Illsley from Omdia told us on the podcast that hyperscalers would inevitably turn to nuclear because renewables alone aren’t enough? He called it perfectly: Google is spending $75 billion this year on data center capacity and betting $9 billion on nuclear, while Microsoft will restart Three Mile Island and Amazon continues to expand its partnership with Talen Energy. The stakes couldn’t be higher for telcos trying to carve out their piece of the AI infrastructure market—your competitors are literally building nuclear reactors to ensure they have the power foundation needed to win the next decade of computing. If you’re still exploring GPUs-as-a-Service strategies, can you compete at this scale?
The World Economic Forum’s Future of Jobs Report dropped some excellent reality on the AI job panic: by 2030, AI will eliminate 92 million jobs but will create 170 million new ones—a net gain of 78 million jobs globally. Based on data from over 1,000 companies across 22 industries, the report shows 77% of employers plan to upskill workers while 41% will reduce headcount as AI automates tasks. But here’s the kicker: 63% cite skills gaps as their biggest barrier to transformation. AI isn’t stealing jobs; you’re not skilling up your workforce fast enough. The operators who thrive in the next decade won’t be the ones who resisted AI and automation; they’ll be those who retrained their people for the 170 million new roles that AI created, starting NOW.
Forward-thinking developers are mastering “compounding engineering,” or building AI systems that learn from every code review and bug fix. Feature delivery goes from weeks to days when teams use AI agents that automatically apply lessons from previous pull requests, creating self-improving development pipelines that get exponentially faster over time. Meanwhile, your legacy vendors are still relying on old-school coding methods to deliver your change requests. Telco leaders: are you ready to adopt the emerging, new school vendors? Or will you keep burning money (like the $1.2 billion a year AT&T pays Amdocs) on vendors that trap you in yesterday’s approach? 🧐
DR, network engineer? This past weekend my home internet went out, and with the US Open tennis tournament starting, you can understand why this was an emergency! I have a complicated Ubiquiti setup to support my 2 gigabit lines to my house. I started by blaming the ISPs (sorry AT&T and Spectrum!), then used ChatGPT-5 to troubleshoot by taking photos of my equipment and UniFi console. Final RCA: a bad gateway AND a failing hard drive. ChatGPT walked me through the troubleshooting, and helped me order new equipment. Next stop: I'm ready to climb a cell tower armed with ChatGPT on my iPhone to fix your RF optimization issues! AI FTW! 🦾
