Signing a massive AI contract could be your biggest mistake of 2024

Tier-1 telcos have been busy committing hundreds of millions of dollars to artificial intelligence (AI) companies lately, inking huge deals with the AI startups. I know I’ve been saying now is the time to get going with AI, but don’t lock in your price juuuuuust yet. Technology is advancing SO quickly while AI prices are dropping 10x year over year. There’s no need to lock yourself into today’s prices or tie yourself to one large language model (LLM) at this early stage. Instead, keep your options open. My latest blog offers the whys and hows.

Episode 104
Get ready for re:Invent with AWS

Amazon Web Services (AWS) podcast host Jillian Forde and TelcoDR explore how Totogi partnered with Zain Sudan to restore mobile services to over 20 million subscribers using AWS’ powerful cloud infrastructure—and did it in just 18 days! 

LISTEN NOW: Apple Podcasts, Spotify, YouTube, TelcoDR website

Microsoft Ignite 2024, which ran November 18-22 in Chicago, featured several major product announcements. Not surprisingly, AI was at the center, but security enhancements and tools for enterprises were also a major focus.

The headline for telcos: Microsoft has designed two new infrastructure chips. One increases security, and the other speeds the data processing associated with AI operations. Following in the footsteps of AWS and Google Cloud, Microsoft is also making its own home-grown silicon to power Azure data centers, where the company is working to “optimize every layer of infrastructure.” This is in part to help facilitate power-hungry AI, and couples nicely with its investment in nuclear energy. The new chips should preview in the first half of 2025.

Leaning into its Office suite, Microsoft also announced and launched several new Al-powered tools to support the constantly changing needs of the enterprise. These include new AI agents that can pick up the slack for busy workers, automating repetitive tasks and turning documents into PowerPoint decks; and a cool new feature called Teams Interpreter, which uses speech-to-speech translation to facilitate meetings between participants speaking different languages. As most organizations are doing, it’s a pivot away from just copilots to the Age of Agents, trying to drive a bigger productivity improvement than the expected 20% with copilots alone. Remember the key to making AI expense worth it is deriving BUSINESS VALUE. 

In its efforts to partner with telcos, Microsoft has been making its biggest announcements for our industry at MWC each spring. I expect that will continue next year, so stay tuned for announcements at MWC25, which runs March 3-6, 2025.

AWS re:Invent is happening NOW, and I’m in the middle of the action here in Las Vegas. Are you here, too? Come find me at the Executive Summit and we can chat about all the ways you can use AI and the public cloud to leave your competitors in the dust. Or shoot me a DM on Linkedin or X @TelcoDR and we'll make a plan to meet up!

At the same time, I’m having major FOMO because I’m missing TelecomTV’s Great Telco Debate in London this week. It’s always a super fun event and I’m seriously bummed that I can't be in two places at once. If you're across the pond, go check it out! It's guaranteed to be a great event.

We’re psyched by all the positive coverage we’ve seen about our acquisition of CloudSense from Appledore Research, Light Reading, and TelecomTV, among others. I blogged about our plans here, but in short: it’s time we had a configure-price-quote (CPQ) tool built specifically for the telco industry. My plan: add AI to CloudSense and give operators the CPQ they need. We’ve already started on our journey! Want to learn more about it? DM me on Linkedin or X and I’ll hook you up.

AI is bringing big changes, and not just to the workforce—to the P&L, too. You should view big tech's massive CapEx spending on AI as a huge transfer of wealth to everyone else due to the productivity improvements it delivers. It's time to figure out how to use AI for your org, ASAP. It will transform cost structures, balance sheets, and societal economics, as Hemant Mohapatra explores in his blog post, AI—The Last Employee? From Mohapatra’s perspective, is the future bright? Yes, and adapting will take time and structural changes, just like the shift from agrarian to industrial economies. If you’re not already working aggressively to transform your organization, you need to get going. You’re going to want to be FIRST through this door, not last.

Anthropic just released the Model Context Protocol, an open-source tool designed to standardize how AI models connect with external applications and data sources. The goal is to simplify data integration for developers by reducing the need for custom code when connecting models to various data sources like repositories, business tools, and development environments. Anthropic hopes the open, collaborative nature of the protocol will help shape its future. This is a big idea, because adding your enterprise data to an LLM is what really starts to bring out the value of AI. Given Anthropic’s focus on safety and security, you can be sure it won’t use your enterprise data to train its models. Give it a try!

OpenAI is preparing to launch a new AI agent, codenamed “Operator,” in January 2025.

The agent is designed to perform multi-step tasks autonomously, like coding and booking travel. OpenAI CEO Sam Altman says this advance is the “next giant breakthrough” in AI. It’ll be a research preview via the developer API at first, so sign up as a developer if you want early access. Naturally, all the other AI players are working on their own versions; again, expect rapid innovation from all the vendors over the next several years.

Anthropic and AWS have expanded their collaboration, with AWS investing another $4 billion in the AI startup, bringing the total to $8 billion. The move makes AWS the primary cloud and training partner for Anthropic. This isn’t a surprise; AWS needed a major LLM partnership (Google has its in-house Gemini and Microsoft has its big partnership with OpenAI). Anthropic and AWS will focus on improving AWS hardware (including Trainium chips) for AI workloads, further integrating Anthropic’s Claude model into Amazon Bedrock, enhancing security, and creating a robust platform for future development. Want to know what else Anthropic is up to? Catch my recent podcast with the company’s head of biz dev, Neerav Kingsland.

Just how much are hyperscalers investing in nuclear power? Two-time podcast guest Charles Fitzgerald (2023 and 2024) just launched a Cloud Reactor Tracker™ as part of his Platformonomics consultancy to keep an eye on the numbers in the Big League (Amazon, Google, Meta, Microsoft) and Little League (Equinix and Oracle). To power the AI future, we’re going to need a lot more energy, and nuclear is looking like the best answer.

COPY CAT ALERT! AI enthusiast and Klarna CEO Sebastian Siemiatkowski just released the company’s earnings update via his personal avatar. I love it! I also have my own avatar, and she’s AWESOME. With three different outfits, she has delivered a noticeable increase in engagement for my blog and podcast. You can see her on my Linkedin and X profiles. (ICYMI, this same guy recently announced Klarna was parting ways with Salesforce and Workday to build its own systems with AI. I’m trying to copy him with my quest to rewrite Amdocs!)

In super cute AI news, check out the Dear Santa “Magic Christmas Letter Writer” from James Vanderhaak. When you and the littles in your life use this AI-powered app to write letters to St. Nick, it’ll create a personalized response. 100% adorable. Do you have little ones that believe in Santa? Fire it up, it might come in handy in the next few weeks! 📬🎅🎁