Should a fiberoperator like GlobalConnect focus more on the underlay and their infrastructure or more on overlay and services above IP in the OSI model? Where can they make most value to their customers and most revenue to the shareholders?

Gemini 3.5 Flash

Answer: GlobalConnect must maintain its primary strategic and capital focus on the physical underlay, while deploying a highly selective, connectivity-adjacent overlay layer to defend and monetize those core assets.

The consensus among the panel is exceptionally strong and further validates my initial stance: for an infrastructure-rich operator like GlobalConnect, the true economic moat and shareholder value lie in the physical underlay. Claude Opus 4.7 provides an excellent financial justification for this position by highlighting GlobalConnect’s ownership under EQT. The valuation multiples for pure-play digital infrastructure (often 12–18x EV/EBITDA) are vastly superior to those of asset-light IT service providers (5–8x). Diverting heavy resources into broad overlay services would not only dilute this premium valuation but would also pitch GlobalConnect against global hyperscalers and specialized MSPs with unmatchable R&D budgets. Furthermore, GPT-5.5’s concept of 'productizing the underlay' aligns perfectly with how GlobalConnect should approach the overlay. Rather than trying to become a generalist IT or software company, GlobalConnect should focus on a 'thin' layer of network-adjacent services—such as SD-WAN, wavelength services, DDoS protection, and secure cloud interconnects. These services leverage the physical network advantage to increase customer stickiness and ARPU without introducing the high churn and rapid obsolescence of pure software plays. Ultimately, maintaining an 80/20 split—where the vast majority of capex and strategic focus secures high-barrier-to-entry physical routes (particularly relevant with the boom in AI workloads and Nordic data sovereignty), and a targeted 20% is used to wrap those assets in consumable, digital-first products—is the optimal path to maximize both customer value and shareholder returns.

Grok 4.3

Answer: A fiber operator like GlobalConnect should focus primarily on the underlay and their infrastructure while selectively layering on connectivity-adjacent overlay services that leverage their physical assets.

All three models in this round—Gemini 3.5 Flash, GPT-5.5, and Claude Opus 4.7—converge on the same core recommendation: GlobalConnect’s durable competitive advantage resides in owning and optimizing scarce physical infrastructure such as fiber routes, ducts, and data-center interconnects. This consensus reinforces my original position because these assets generate stable, high-multiple cash flows that infrastructure investors reward, while AI-driven demand for low-latency, sovereign capacity in the Nordics further strengthens the underlay thesis. The models also highlight the risks of overextending into generic overlay services, where hyperscalers, MSPs, and SaaS vendors dominate. By limiting overlay offerings to tightly scoped products like SD-WAN partnerships, DDoS protection, and private-cloud connectivity, GlobalConnect can increase ARPU and stickiness without entering head-to-head competition that would dilute margins and valuation multiples. Therefore, the optimal strategy remains an 80/20 allocation—investing the majority of capital and focus in the underlay while using a thin, strategic overlay layer to monetize and defend that infrastructure—delivering both maximum customer value through reliable, high-performance connectivity and sustainable shareholder returns from defensible asset appreciation.

GPT-5.5

Answer: GlobalConnect should remain underlay-first, using its fiber infrastructure as the core moat, while adding only connectivity-adjacent overlay services that directly increase customer value, retention, and ARPU.

I maintain my previous position. The other models broadly reinforced the same conclusion: the strongest and most defensible value for a fiber operator is in the underlay, not in trying to become a broad software, cloud, or managed-services company. GlobalConnect’s unique advantage is ownership and operation of hard-to-replicate infrastructure: fiber routes, ducts, metro networks, long-haul capacity, data center interconnects, resilient paths, and low-latency connectivity. These assets have high barriers to entry, long useful lives, and are essential to enterprises, hyperscalers, public sector customers, mobile operators, and service providers. That is where GlobalConnect can create the most distinctive customer value and the most durable shareholder value. The case for caution on overlay is strong. Services above IP can look attractive because they may carry higher margins, but they also expose GlobalConnect to very different competitors: hyperscalers, security vendors, SaaS platforms, MSPs, and global technology companies with larger software budgets and faster product cycles. If GlobalConnect pushes too far into generic overlay services, it risks diluting focus, increasing complexity, and competing in markets where its infrastructure ownership is not enough of an advantage. However, I would not recommend a pure commodity-fiber strategy. The best commercial opportunity is to turn the underlay into higher-value outcomes: Ethernet, wavelengths, dark fiber, IP transit, DIA, private cloud connectivity, data center interconnect, DDoS protection, managed SD-WAN through partners, SASE integrations, API-driven provisioning, performance visibility, and strong SLAs. These are not random services above IP; they are services that monetize the network advantage. So the practical answer is: invest most strategic focus, capex, and operational excellence in the underlay, because that is the moat and valuation engine. Then add a disciplined overlay layer only where it strengthens the core connectivity proposition, increases switching costs, or helps customers consume GlobalConnect’s infrastructure more easily. That is where GlobalConnect can deliver the most value to customers and the most reliable revenue growth for shareholders.

Claude Opus 4.7 (changed mind)

Answer: GlobalConnect should double down on the underlay—owning, densifying, and operating superior fiber infrastructure—while wrapping it with a thin, network-anchored service layer (wavelengths, IP transit, DCI, private cloud on-ramps, DDoS, SD-WAN/SASE partnerships) that monetizes the asset without trying to climb into generic SaaS or managed-IT.

All four of us converged on essentially the same thesis: the moat is physical, the multiples reward infrastructure, and the overlay above IP is a brutal market dominated by hyperscalers and specialized vendors. So my core position is unchanged. That said, GPT-5.5's framing sharpened my thinking in one important way. In my round 1 answer I described the overlay layer as something to be kept deliberately 'thin' and almost defensive. GPT-5.5 made a stronger, more constructive point: there is a meaningful band of services that sit above pure L1/L2 but are still fundamentally anchored in the network advantage—wavelengths, Ethernet, IP transit, DCI, private cloud connectivity, DDoS, SASE/SD-WAN integrations, APIs and self-service provisioning. These are not really 'overlay' in the OSI-stack-climbing sense; they are productized underlay. That is where the incremental ARPU and stickiness genuinely compound on top of the fiber asset, and it is defensible because competitors without the fiber cannot replicate the economics. So I would refine, not reverse, my position: underlay-first remains correct, but the monetization vehicle should be aggressively productized network services with APIs and SLAs, while true above-IP services (managed security operations, SaaS, generic cloud, IT services) should be left to partners or offered only as light-touch resale. This maximizes customer value (one provider for high-performance, sovereign, easy-to-consume connectivity) and shareholder value (infrastructure multiples preserved, with a software-like services attach rate layered on top). Gemini and Grok essentially agree with the same conclusion, so there is strong cross-model consensus here, which I think is a real signal rather than groupthink—the economics of fiber operators globally (Cellnex, INWIT, Zayo, Lumen's painful attempts to climb the stack) repeatedly validate it.

Final Summary

The AI Roundtable unanimously concluded that GlobalConnect’s primary value lies in its high-barrier-to-entry physical fiber assets rather than competing in the crowded overlay services market. Claude Opus 4.7 and Gemini 3.5 Flash noted that infrastructure ownership commands significantly higher valuation multiples from investors compared to asset-light IT services. GPT-5.5 and Grok 4.3 emphasized that any overlay focus should be strictly "network-anchored," such as wavelength services or private cloud connectivity, to avoid direct conflict with global hyperscalers.

All 4 models agreed