Networks, Resilience & the SpaceX Reality Check | Ep2

Intelligence & Noise is hosted by Peter Jarich and Tim Hatt and comes out every week on YouTube and all major podcast platforms.

Spotify: https://open.spotify.com/show/033rTmN6BootY1WNYX5oPe 
YouTube: https://www.youtube.com/@IntelligenceGSMA 
Apple: https://podcasts.apple.com/us/podcast/intelligence-and-noise/id1896878144 
Amazon Music: https://music.amazon.co.uk/podcasts/06164504-caf1-454b-a46d-3904f7418a20/intelligence-and-noise 

Intelligence & Noise \ Episode 2 summary

The second episode of Intelligence & Noise takes a theme that runs quietly through most weeks in mobile — resilience — and makes it explicit. From disaster warning systems in Somalia to the SpaceX IPO prospectus, the thread connecting this week's stories is the same question: what are networks actually for, and what happens when the numbers don't add up?

Co-hosts Peter Jarich and Tim Hatt draw on GSMA Intelligence research and direct operator conversations to cut through a week that had no shortage of big claims and bigger headlines.

Resilience as infrastructure, not afterthought (03:02)

Two stories from the same week point to the same thing. A $13 million early warning system going into Somalia. A group of Japanese operators investing in portable battery backups to keep networks alive through extreme weather events. Neither story made the top of the news cycle. Both matter more than most things that did.

Peter and Tim read these not as isolated announcements but as a signal of where the industry is heading. Somalia is a mobile-first country — mobile banking, mobile government services, mobile internet. When a network goes down after a disaster, the consequences are not an inconvenience. They are the difference between people being reached or not. Tim makes the point that these are no longer edge cases. Typhoons, hurricanes, extreme weather events — these are the new normal, and network preparedness needs to be treated accordingly.

6G, ISAC and the sensing opportunity nobody is explaining clearly (09:50)

One of the genuinely new things coming out of 6G is integrated sensing and communications — ISAC. Peter walks through what it actually means. The idea is straightforward: use the signals that mobile networks already emit for communications to do something more. To sense the physical world. Traffic flow. Object detection. Drone identification. Early warning.

The use cases range from the mundane to the strategically significant, and the conversation touches on where the technology works well, where the spectrum and cost constraints bite, and why sensor fusion — pairing mobile sensing with cameras, satellites and other inputs — is likely where the real capability emerges. The connection back to Somalia is direct. A network that can warn people about an incoming disaster is only useful if it stays up when that disaster arrives. End-to-end resilience is the point.

The SpaceX IPO and a $28.5 trillion question (15:52)

GSMA Intelligence published a spotlight on the SpaceX IPO prospectus. The prompt was simple — members, customers and colleagues were all asking the same thing. The numbers in the filing are very large. What do they actually mean?

The headline TAM figure is $28.5 trillion. Most of that sits in enterprise AI applications, which Tim and Peter acknowledge is difficult to model and connected to SpaceX's longer-term vision of compute infrastructure beyond Earth. The part the team focused on was Starlink Mobile — the direct-to-device connectivity business where SpaceX has positioned itself as a provider for everyone, everywhere.

When GSMA Intelligence ran the numbers against what the technology can actually deliver — applying realistic throughput thresholds, available spectrum and capacity constraints — the serviceable market for Starlink Mobile came out at around 5% of SpaceX's stated figure. The coffee analogy Peter reaches for is useful. A total addressable market can legitimately mean everyone who could theoretically buy your product. What matters for investors is how much of that you can actually serve. The gap between those two numbers is where the real conversation is.

The team is careful to be clear on what this is and isn't. SpaceX has a genuine first-mover advantage in low-earth orbit satellite connectivity. Starlink has real momentum. The point is not to dismiss the business. It is to apply the same evidence-based rigour that GSMA Intelligence applies to any market claim.

The World Cup as a connectivity test bed (25:25)

The FIFA World Cup arrives this week and Peter has been doing analysis on what it means for operators. The surface story is familiar — networks upgrade capacity, data records get broken, telcos issue press releases. The more interesting story is underneath that.

What operators are actually deploying at this tournament goes well beyond fan connectivity. Network slicing for broadcasters. Roaming and speed-tier services for international visitors. T-Mobile's live translation service, launched from the core network and timed, probably not coincidentally, for an influx of non-English speaking fans. These are B2B and B2B2C revenue stories, not just infrastructure stories. And with generative AI now embedded in how fans consume and share content, the uplink data picture at this tournament will tell the industry something new about what demand actually looks like when AI is part of the experience.

Peter and Tim also find time to debate who exits the group stage first — the US or Canada.

OpenAI, agents and the profitability question (31:12)

An FT story from the weekend sets up the final main topic. OpenAI, approaching a likely IPO later this year, is signalling a strategic pivot towards agentic AI and enterprise revenue. The consumer chatbot is closing in on a billion users. The business model question is whether that translates into the kind of sustainable revenue that investors will want to see at IPO.

Tim and Peter note the contrast with Anthropic, which has gone heads-down on B2B and coding — a direction that is showing up in its revenue trajectory. The open question for OpenAI is whether the personal agent framing is a genuine strategic commitment or a holding position. The consumer agentic market, Peter argues, may be more overhyped than it appears. The evidence from people buying Mac Minis to run local AI models and then giving up is one data point. The broader question of whether consumers will actually pay for agentic capability at scale is still unanswered.

On the road (36:44)

Tim reports from an ESG workshop in Brussels with GSMA's European team, European regulatory authorities and a group of mobile operators. The notable development: biodiversity monitoring is emerging as a meaningful strand of the telco ESG story. Sensing capabilities — the same capabilities discussed in the ISAC conversation earlier in the episode — are being embedded on towers and mobile infrastructure to track environmental indicators. It is not a headline story yet. Tim thinks it will be.

Coming up, the co-hosts flag MWC Shanghai and DTW Ignite as the next stops on the road.