TelcoDR’s Summer Reading List 2026
The core problem: The first half of 2026 produced more strategic AI thinking than most telco leadership teams had time to absorb: Benedict Evans on the platform shift, Alex Karp and Satya Nadella fighting over who captures AI value, and a pricing war among the frontier labs. TelcoDR’s summer reading list collects the essential posts, podcasts, and frameworks so operators can catch up before the fall planning cycle.
Half your team is returning out-of-office replies right now, but the AI industry doesn’t quit. In the past five weeks alone, Alex Karp went on live TV to torch token pricing, and Satya Nadella coined a whole new economic paradox about who owns the knowledge you feed your AI vendor. The strategic ground is shifting in the exact months telco is least likely to be watching.
My suggestion: use the quiet to your advantage. Get yourself a cold drink, set up a chair in front of the A/C, and get ready to be the most well-informed person in the company once your colleagues get back from vacation.
Every summer, we like to look back over the year and shine a spotlight on the best blog posts and podcast episodes we’ve published. And this year, for the first time, we’re also taking a look at some remarkable thought leadership from people besides me! Because hey, I’m not the only one with good ideas, right?
Read it now. Thank me later. Here’s the list!
AI thought leadership
The last six months have given us true thought leadership when it comes to AI. If you missed these nuggets the first time around, now’s your chance to catch up. If you already read them, you may want to read them again. They’re that good.
Benedict Evans’ AI eats the world
Analyst Benedict Evans gave a 16-minute keynote at MWC26, talking through his latest presentation, “AI eats the world.” In it, he framed AI as the next platform shift after web and mobile, with one implication operators keep underselling: AI dramatically lowers the cost of building software. When AI can generate, integrate, and deploy code at a fraction of the old cost, the $50 billion+ professional services industry built on BSS complexity starts to look unnecessary.
The updated version of AI eats the world, from May, argues that AI is becoming a general-purpose infrastructure layer, like cloud or electricity, but its lasting value will likely be above the models—in apps, workflows, and product design. The winners will be those who turn raw model capability into useful, sticky, scalable software that people actually adopt. For telcos, that means deploying AI to production—not admiring the platform shift from the sidelines.
Alex Karp’s broadside on OpenAI and Anthropic
On July 1, Palantir CEO Alex Karp made an important complaint on Squawk Box: enterprises are paying for tokens while the frontier labs capture the value, and the real money sits in the platform layer that turns context into action. Telco leaders, take a cue from Palantir’s nine-point framework about AI sovereignty and ask your AI vendors: who keeps the data? Who controls the weights? Who owns the alpha? If they squirm, you have your answer.
Satya Nadella’s X articles on AI architecture
In June, before Karp’s polemic, Microsoft CEO Satya Nadella posted an article on X, A frontier without an ecosystem is not stable, putting a stake in the ground about where AI value lives: in the learning loop on top of the model—what he calls “token capital.” Can you swap your underlying model without losing your “company veteran” expertise? If not, you don’t own your IP. You’re renting it. And don’t expect a model vendor to voluntarily change that for you. The Totogi Ontology is exactly the architecture Satya describes, built specifically for telco.
A month later, on July 12, Nadella followed up with The Reverse Information Paradox. On the surface, his point is that to make AI useful, companies have to reveal proprietary knowledge to the model and provide corrections to it, which the model uses to improve itself. You pay for intelligence twice: once with money, and again with the know-how you hand over to make it useful. And yet, users are contractually barred from building products or services that compete with the models based on what they learn by using them. If learning flows in only one direction, he argues, the economic value flows with it—toward whoever owns the learning infrastructure, not whoever created the knowledge.
Well, this isn’t neutral economic philosophy. It’s (rather elegant) whining from someone who’s now on the back foot in the AI race. Anthropic just passed OpenAI in enterprise revenue and is collecting knowledge from its 42% share of the enterprise coding market (to OpenAI’s 21%). Nadella is asking the winning lab to share the wealth from a position of relative weakness rather than pure principle.
Telco, your position is even weaker than Microsoft’s. If Nadella can’t force reciprocity out of a vendor, neither can you. A telco’s token capital lives in an operational ontology—the encoded logic of how your company actually works: billing rules, provisioning states, eligibility constraints. A layer YOU SHOULD OWN that gets smarter with every churn save, every order, every network fault—and one that survives any model swap underneath.
AI-Native Telco Accelerator Index
This one hasn’t landed yet, but circle September 8-9 on your calendar. That’s when the AI-Native Telco Accelerator (ANTA) releases its first AI-Native Telco Index at its forum in Düsseldorf: a ranking of 50 Tier 1 and Tier 2 operators against 20 markers of AI-native transformation, from foundational readiness and strategic intent to execution evidence and transformation velocity. For the first time, you’ll be able to see exactly where you stand against your regional peers—and the industry will see who’s actually deploying versus who’s still admiring the problem.
I’m proud to say Totogi is one of ANTA’s two founding technology partners, working alongside a steering board of Axiata, Deutsche Telekom, Orange, NTT Docomo, and Rakuten Mobile, plus the GSMA. We joined because moving this industry from scattered pilots to scaled deployment requires shared benchmarks and collective action. The last time I was in Düsseldorf, we brought an AI engineer and built a BSS in nine hours. Seems like the right city to start keeping score. See you there.
DR’s best blog posts
AI is changing so much of how we operate in telco—and we’re just getting started. Now that the initial hype phase is over, it’s time to get seriously strategic about how we rebuild this industry for the future. The following posts all take a step back to ask larger questions about how best to invest in and incorporate this technology.
The Big O of AI
What does it cost to run an AI application at scale? Nobody knows. Not even the AI labs, which is why they just went to war over pricing. Fifty years ago, computer science figured out how to size a design’s cost before running it: Big O notation. We need one for AI. I took a stab at the notation in the hopes that it’ll change how we think about AI app design.
Own your learning loop
Should telco own its learning loop, or rent it? It’s the biggest infrastructure decision you’ll make this decade. Put AI decisions in the foundation model, and you’re signing up for (more) decades of BSS vendor lock-in. If you want to break free, the architectural fix is to separate the decisions from the model in an ontology layer you own.
SKT’s stake
In August 2023, SK Telecom made a $100 million investment in Anthropic. Now, SKT’s stake is worth $1–2.6 billion, prompting Morningstar to reclassify it as an AI stock, expanding its P/E from 11x to 60x. In this post, I break down what SKT actually did, what everybody else is doing instead, and why the market knows the difference.
Where do decisions live?
Becoming AI-native requires you to answer an important question: where do decisions live? Legacy vendors are telling you to put decisions in AI agents. It sounds like transformation, but it’s not. It’s the same broken logic, just faster. An operational ontology—like the Totogi Ontology—gives you the control you need at the speed of AI.
Nail your AI architecture
AI finally gives telcos a way out of paying for every change to their BSS/OSS systems. But dinosaur vendors are trying to maintain the status quo by pitching agentic AI with business logic baked in. That’s the wrong answer. This blog outlines why agentic AI without an ontology will keep you where you are now: trapped, paying by the change.
Podcasts you may have missed
I love getting the inside scoop from telco industry leaders, talking one-to-one on my Telco in 20 podcast. The following episodes stand out because they feature pioneering companies doing innovative work, sometimes swimming against the current of an industry intent on going the other way.
AT&T bets on the last mile (Shawn Hakl)
AT&T’s AWS Interconnect – last mile is fiber and fixed wireless that plugs directly into Amazon Web Services and is engineered for AI workloads. It’s a zig when other operators are zagging in the race to claim a piece of the AI infrastructure stack. Listen in as Shawn Hakl, SVP of product at AT&T Business, and I talk about partnering with hyperscalers (instead of competing).
Swisscom is building the AI-first telco (Mark Düsener)
Telcos are talking the talk about becoming AI-first. Swisscom is walking the walk, doing a ground-up rebuild of the software core behind its consumer and enterprise business units. Mark Düsener, chief technology and information officer at Swisscom, and I unpack how the operator is doing it, and what it’s learning in the process.
Microsoft’s framework for autonomous networks (Rick Lievano)
Buried in all the AI hype, there are real deployments with measurable impact. In this episode, I talk with Rick Lievano, Microsoft’s CTO for the Worldwide Telecommunications Industry, about Redmond’s vision for network autonomy, actual customers achieving breakthrough results, and the barriers still holding telcos back from deploying AI agents at scale.
Show me the money (as a video or a podcast)
Considering an AI investment? Ask your vendor the all-important question: can you show me the money? When it comes to actual return on investment, most can’t prove business benefit. This video, a version of the speech I gave at this year’s MWC as Totogi’s CEO, highlights how the Totogi Ontology has delivered real value to Zain, shrinking, then eradicating, an expensive dormant cell problem.
So that’s your summer reading list—proof that even in the slowest month of the year, telco isn’t standing still. AI is moving fast, and the operators paying attention now are the ones that will have a leg up in the years to come. Enjoy the sun, enjoy the quiet, and enjoy getting a little smarter while everyone else is out of office. Later, nerds!
Recent Posts
Get my FREE insider newsletter, delivered every two weeks, with curated content to help telco execs across the globe move to the public cloud.
Get started
Contact Totogi today to start your cloud and AI journey and achieve up to 80% lower TCO and 20% higher ARPU.
Transform
A leading mobile network operator reduced alarm noise by 97% and accelerated the resolution of complex network failures with the Totogi Ontology.
Engage
Set up a meeting to learn how the Totogi platform can unify and supercharge your legacy systems, kickstarting your AI-first transition.
Understand
The term “ontology” has become mainstream—but everyone saying it isn’t talking about the same technology. This blog explains what a real ontology does.
Frequently Asked Questions
It’s the operational logic that makes your AI actually useful: the billing rules, provisioning states, and eligibility constraints that encode how your company really works. Satya Nadella calls it “token capital” in his June X post. The point is simple: if you would lose your “company veteran” expertise if you swapped out your underlying model, you don’t own your IP. You’re giving it to someone else, and then renting it back. That’s why we built the Totogi Ontology, so telcos own that layer instead of leasing it back from a BSS vendor forever.
No, and any vendor telling you otherwise is selling you the same old lock-in with an AI coat of paint. Decisions need to live in an operational ontology you control, separate from the model itself. Bake business logic into the agent and you’re still paying by the change, just with extra steps. That’s the whole argument behind our “Where do decisions live?” and “Nail your AI architecture” blogs.
It’s real, but uneven. Nvidia’s 2026 State of AI in Telecommunications report found that 90% of respondents say AI is increasing revenue and cutting costs, with the biggest wins in autonomous networks (50%), customer service (41%), and internal process optimization (33%). AI is working for our industry, and the telcos that adopt it early are giving themselves a leg up that will pay off later. Don’t get stuck in laggard mode.
It proved that betting on AI infrastructure early pays off—literally. SKT put $100 million into Anthropic back in August 2023, and that stake is now worth somewhere between $1–$2.6 billion, enough that Morningstar reclassified SKT as an AI stock and expanded its P/E from 11x to 60x. The market isn’t rewarding telcos for talking about AI.
Steal Palantir CEO Alex Karp’s framework: who keeps the data, who controls the weights, and who owns the alpha? If your vendor squirms answering any of those, that’s your answer. Enterprises are paying for tokens while the frontier labs and platform layers quietly capture the actual value. Nadella, Karp, TelcoDR: we’re all talking about the same thing.