DJ Khanmamedov, MBA, PMP, CSM

For those of you who following the messaging industry, many of you know that I have gone from enthusiastic to skeptical back to enthusiastic about RCS. Now that iOS 18 is supporting RCS and we have interoperability of RCS between Android and Apple, I'm back on the enthusiastic bandwagon. Especially, once RCS Business Messaging (or RBM) becomes more available on iOS. For me, that's not the case yet (USA, AT&T). But there is a new wrinkle, and I have questions. I hope this community can help address these, so we can have a better view of the implications. I'm talking about what is in iOS 18.2 (currently in Public Beta). Similar to Android, certain apps may now specified as "default" and this includes messaging apps. So, for the first time, users may indicate a different app, other than Apple Messages (or iMessage) as the default messaging app. A Forbes article recently published an article about it and the author said it was "bad news for Google" (and as such, for RCS). OK... we've just gotten here to RCS bliss, so now WHAT? Here's the article: https://lnkd.in/eHe5Kr_C Now, my questions: 1) I don't have access to the Apple Developer's program - does Apple make iMessage chat APIs? What if one user is using the default Apple Messages and the other an alternative? Both are iOS users. Does the alternative (and now default) messaging app have the ability to support "iMessage" chat? Or does that alternative user only able to send/receive SMS/MMS to other Apple users? Or put more succinctly: What does this do to Apple-to-Apple texting when one party is on iOS but not using Apple Messages? 2) The Forbes article implies that any alternative Apple messaging app will not support RCS (or cannot support RCS), thus eliminating one iOS RCS target.... or "bad news for Google." But does it HAVE to be? Whose limitation is this? Google's? Why then, on Android, do alternate apps like Smart Messages seemingly support P2P RCS & RBM (as well as other, RCS-capable messages such as Verizon Messages [since recently shut down] or Samsung Messages [also ended] supposedly support Universal Profile based RCS)? I've heard some limitations for a messaging app (or a 2nd-line messaging app) being unable to interoperate with other RCS messaging apps - even if such an app supports UP RCS and is backed by an IMS-core, using NNI to interconnect. It may be supported, but not without certain conditions. It seems to me that if such alternative apps for iOS 18.2 & beyond were to support RCS, then they should interoperate with other devices running UP-based RCS (like Android Messages or Apple Messages or even another alternative app with RCS support). This does not have to be a threat to RCS as a replacement for SMS/MMS - such an app should simply support RCS.

27

12 Comments

In a groundbreaking study, researchers explored how AI can identify human emotions in online posts, particularly on platforms like X (formerly Twitter). 🧐 By using a sophisticated transformer transfer learning model, they were able to detect specific emotions such as joy, sadness, and anger expressed in tweets with an impressive 84% accuracy. The study revealed that these emotions could directly influence actions, such as driving donations to non-profit organizations, indicating the profound impact of emotional resonance in online communication. As AI technology continues to advance, ethical considerations come to the forefront. With the ability to read and interpret human emotions, it becomes crucial to ensure responsible and ethical use of this capability. Questions surrounding privacy and data protection in the context of emotional detection by AI must be thoroughly addressed as this technology evolves and becomes more integrated into our daily lives. The potential applications of AI's emotional detection capabilities span across various sectors, including marketing, education, and healthcare. However, the responsible and ethical use of this technology is paramount. How can we ensure that AI's emotional detection is used ethically and responsibly, while also safeguarding individuals' privacy? #AI #Emotions #Ethics #Privacy #TechEvolution

9

I found this at the top of my Tweets: Andrej Karpathy's GPT2 video lecture series is running on YouTube. Coverage is comprehensive - first, we build the GPT-2 network - then we optimize it to train very fast - then, we set up the training run optimization and hyperparameters by referencing GPT-2 and GPT-3 papers - then we bring up model evaluation and Please check out the links below https://lnkd.in/dSMW4DsZ https://lnkd.in/dQXuB8iM https://lnkd.in/dubKY3an #gpt2 #llms #generativeai #ml

2

I am constantly amazed by the incredible work led by network innovators and their innovations that drive so much of the world we live in. I think I probably am generating or consuming data almost every second of every day (+sleep tracking). Now multiply that by an astounding almost unimaginable number, and you have the amount of data that is collected, moved and studied by the research community. A community that is driven to expand humanity's horizons and make our world a better place. This article by BBC puts a spotlight on CERN and the fascinating work involving the Large Hadron Collider (LHC) and the network that supports its mission. I hope you enjoy it, and think about it the next time you hit "send." 😊 https://lnkd.in/eqrvMj-D

10

Anthropic just released an update for Claude 3.5 Sonnet (medium model) and 3.5 Haiku (small model) today. 3.5 Sonnet + Cursor is the magic combination for LLM-assisted coding. According to their benchmark, 3.5 Haiku is better at coding than Sonnet. That would be a big deal considering how cheap and fast Haiku is. They also introduced "Computer Use," where the LLM takes control of your desktop. This will be very helpful for useful agents. There is enough "reasoning" in LLMs that they can do many tedious tasks, but unfortunately, they don't yet have agency on your computer. That's the solution for it. OpenAI bought a company for that a few months ago. https://lnkd.in/dQGiH6sE PS: Where is Opus 3.5?

12

#productfeatures One thing that brings me immense joy in life is talking "wireless". This week's wireless theme has been mostly discussing a simple, scalable, cost effective, utilitarian solution that adds native voice and SMS to a Private 4G/5G network. Something I call #PRV8Calling Now you may ask, how does this fit with the picture below? Well I am glad you asked, let me tell you how #PRV8Calling adds value to your Private Network. 1️⃣ It is the easiest solution that I know of to add native voice and SMS to a Private 4G/5G network 2️⃣ It is the most cost effective solution that I know of to add native voice and SMS to a Private 4G/5G network 3️⃣ It is the most scalable solution that I know of to add native voice and SMS to a Private 4G/5G network 4️⃣ It is the fastest to deploy solution that I know of to add native voice and SMS to a Private 4G/5G network 5️⃣ And most importantly it allows for all of your employees the ability to make and receive voice and SMS natively on their handset Most of the press coverage I see highlights the "fancy" use cases deployed using Private 4G/5G. And who doesn't like "fancy" use cases and solutions? But what if there was a quiet majority of venues and enterprises that could benefit from just having the "boring" use case of quality Coverage - Connectivity - Voice - Data service available to them. I think it is time to have options in the market that fits the #userneeds And that can be done while still providing "fancy use cases to the companies that are ready to deploy them like: Tesla Dallas Fort Worth International Airport (DFW)Toyota North America Purdue University Airport United States Department of Defense Associated British Ports Visit us @ networkfx #networkfx #PRV8Calling #SuperNOVA #PRV8Connect #PRV8Hub #PRV8Steering #offload3.0

17

2 Comments

I was chatting with my friend Claude Seyrat, the CEO of Firecell yesterday about the future of private networks. I have long held the theory that the #privatenetwork market is divided in two categories. 1. Entities who want a connectivity experience that is better than #wifi, e.g. basically a little simple #QoS with prioritization of devices, inexpensive and simple, plug and play. This category is price sensitive and has no desire to understand 3GPP acronyms. 2. Entities who want a connectivity experience than is better than commercial network, either because there is no commercial network available (mining, energy...) or because they need programmable connectivity for their specific use (defence, manufacturing, robotics...). This category will invest in specific skillset to plan, create and manage their connectivity experience. In between these, the question mark is roaming / compatibility / hosting with public networks and the role SIs, network operators can / want to play there. What are your thoughts? #private5g {Core Analysis}

38

10 Comments

META OBJECTIVE Removing such a wide disparity has to be the main objective, not being the 3rd biggest economy or being many trillion economy or achieving double digit GDP growth. This has to be the driving objective. If the main objective is incorrect then all the efforts will drive the country in wrong direction, faster.

3

The "Wow" effect of Voice AI is near the end. What's next? I just saw an incredible demo from Happyrobot—a Voice AI that negotiates prices, clarifies misunderstandings, and manages complex interactions. Impressive? Absolutely. But we're entering a new phase. To level up, we need: ⟡ Improved reliability ⟡ Scalable systems ⟡ Better tool calling and turn-taking ⟡ Advanced end-of-thought detection ⟡ Speech APIs with better language support ⟡ Voice AI that remembers And in 2025, I expect: ↳ Voice AI infrastructure to take center stage ↳ Seamless integration into contact centers ↳ Marketplaces for Voice AI integrations ↳ Built-in storage and function-as-a-service Who's ready for the next chapter of Voice AI? (Congrats to Happyrobot for the killer demo!) ▶ Need help building your Voice AI stack? Let's talk—DMs are open.

46

10 Comments

New CCS Insight blog post published reflecting on announcements at Google I/O and the delicate balancing act Google needs to deliver between itself and Android licensees. "Google is treading by having a more integrated approach to hardware, software and services. We believe the shift toward a more cohesive offering here will concern many Android licensees. In reality, there’s little these companies can do; there are currently no credible large-scale alternatives to Android. Google will argue that it can find a balance, and its continued investment in the Android platform underlines its desire for everyone to be successful, but it needs to be careful to avoid growing tensions with its partners." Free to read here: https://lnkd.in/eR3t_xEg Geoff Blaber, Leo Gebbie, Bola Rotibi

21

2 Comments

"The Intelligence Age" essay by Sam Altman at OpenAI explains how we're able to cross the next tech evolution chasm in 3 words, "deep learning worked", anticipates superintelligence in just a few thousand days, and hints GPT-5's been already trained 🥷 While some founders and entrepreneurs've been writing essays for a while, I keep coming across more and more authors recently. Seems to be a thing and makes sense 🪁 Writing longer forms helps one clarify the thoughts and get the message across in a compelling way. Link to the essay in comments. #ai #llm #comms

7

1 Comment

Good post from VOX SOLUTIONS on Twitter’s $60 Million Scam. There's $100Ms being scammed from web brands using #AIT #SMS. Hence the push to alternative technologies like #passkeys and even good old #email. https://lnkd.in/eCB7ZVXf Daniel Gill from Augnet shared how this scam works and how it can be stopped. AIT (Artificially Inflated Traffic) is being used to destroy the routing tables enterprises (e.g. Google, Meta, Amazon, Twitter) use for SMS; and because data in SMS is sent in the clear its about to become a massive source for fraud including account take over (#ATO). https://lnkd.in/eJAMj4ni There are also processes that can help reduce the opportunity for fraud. Uku Tomikas highlighted if the SMS rates were openly published, the industry can focus on value add. https://lnkd.in/eCSK2SNE. Such a open, honest, and simple step. Robert Vis showed how focusing on the service and bringing the price of SMS close to cost can also shift the business to be customer focused. https://lnkd.in/ebAq2HFk We had an interesting discussion on whether telcos are culpable in the HORISEN podcast, https://lnkd.in/dhE9pdAP. Fabrizio Salanitri showed how telcos can take back control of the SMS experience for their customers. None of this is rocket science, the industry needs to be open and customer focused. The current phrase is "restoring trust." Unfortunately, the companies that created the current situation are unlikely to be trusted again. Trust can only begin through sharing how this circumstance came about and how it can be stopped. Talking shops with 'codes of conduct' will not engender trust. Action is required through the adoption of technologies such as Augnet and HORISEN, coupled with published SMS pricing. We'll be discussing this and more at TADSummit, https://lnkd.in/duqNyHnP. The event where leaders with conscience have the open conversations lacking elsewhere. Infobip, GSMA, netnumber , Aegis Mobile, WMC Global, Vonage, iconectiv, GMS, Tata Communications, Sinch, CTIA, Kaleyra, Cloud Communications Alliance, #PSTN, SignalWire, Von Coalition, TADHack, Mitto, Federal Communications Commission, Federal Trade Commission, YouMail Inc., YouMail Protective Services, Twilio, tyntec, The Campaign Registry, Robert Gerstmann, Sinch, Messente Communications

19

2 Comments

Next week, I, along with thousands of customers and partners, will be heading to Amazon Web Services (AWS) re:Invent in Las Vegas. I’m eager to dive deeper into AWS’s advancements and strategic priorities, particularly in the following areas: AWS’s Progress in AI: I’m looking forward to insights from AWS on its advancements in artificial intelligence, especially with special guest Andrew Ng, a thought leader in deep learning and founder of Landing AI and the AI Fund. Edge Computing Strategy: Understanding AWS’s vision for the next multi-billion-dollar market in edge computing is high on my agenda. Partner-Driven Revenue Growth: Exploring how AWS is empowering its partners to drive revenue with solutions like Mosaic and Trainium. Sustainable Data Center Innovations: Learning more about AWS’s bold strategy to power its data centers with Nuclear Small Modular Reactors (SMRs), backed by a $500M investment. Telco Market Outlook: I plan to meet with Ishwar Parulkar to gain insights into his perspective and AWS’s positioning in the telecommunications market. If you’re attending re:Invent and have a focus on the telco sector, feel free to reach out to schedule a meeting. I have slots Monday - Wednesday. Safe Travels! Gina Sánchez B. Dafna Yanay Brian McManus Jeffrey Dillon Roy Chua Scott Raynovich Brian Partridge Justin van der Lande

17

8 Comments

Not all models are created equal. While Large Language Models (LLMs) offer impressive capabilities, Domain-Specific Models (DSMs) provide a more practical and efficient alternative for enterprises wanting to adopt Generative AI now but have reservations. DSMs are tailored to specific tasks of a workflow, resulting in higher accuracy and lower costs for enterprise use cases compared to LLMs. As Aidan Gomez, CEO of Cohere, noted, "I can speak to the enterprise world, and there, enterprises are super risk averse. They’re always looking for opportunities, but they’re extremely risk averse. That’s the first thing that they’re thinking about." While replacing humans with Artificial General Intelligence (AGI) might be a distant possibility, DSMs can hyper-automate repetitive manual enterprise workflows today. Generative AI in the enterprise requires well-understood use cases, an iterative approach to automation, and clear metrics to measure success. DSMs can hyper-automate up to 90% of repetitive employee tasks, delivering tangible and measurable value today. In summary, DSMs offer cost-effective, highly accurate, and immediate value for automating enterprise workflows. If you have employees doing manual, repeatable workflows, reach out to learn more.

3

1 Comment

🚀 Big Milestone for LabLabee! 🚀 After months of hard work, we’re thrilled to announce that we’ve closed our new funding round of $3.4M, led by the incredible team at Reach Capital, with amazing support from Classera, Naseej's VP of Business Dev Mohammed Husamaddin, and successful founder & investor Cedric Sellin. A special thanks to our existing investors, Brighteye Ventures and e& capital, for reinvesting in this round! This funding will fuel our expansion into the US while continuing our growth in the EMEA region. None of this would be possible without our dedicated LabLabee team, amazing partners, and early believers who have supported us from the beginning. Reflecting on the journey, it’s incredible to think that it all started when Mahfoud and I were students in Algeria, building small projects during our free time that made an impact. Years later, we took different paths—Mahfoud in startup and edtech with Quizzito and me in telecom with companies like Orange, Capgemini, and Mavenir. It was the perfect timing to build an edtech platform for telecoms, making advanced technologies accessible to the entire ecosystem. Today, LabLabee serves great customers like Orange, @Bouygues Telecom, Ooredoo Algerie, Axian Group, and Deutsche Telekom We are extending an invitation to the entire ecosystem, including: - Telcos facing challenges in upskilling and reskilling their teams in emerging technologies like cloud, automation, and AI. We provide practical, scalable solutions that help telcos bridge skill gaps and stay ahead in this rapidly evolving landscape. - Industrials exploring IoT, edge computing, and private networks, who need hands-on training to equip their teams with the knowledge to build and manage these advanced systems effectively. - Systems Integrators (SIs) that require fast, cost-efficient, and operationally focused skilling programs for their consultants to ensure they can deliver cutting-edge solutions and stay competitive in the market. - Vendors searching for a scalable, virtual lab solution to enhance their teams' skills while also empowering their partners and customers. Our platform provides a versatile tool that enables rapid, real-world, technical training for all stakeholders, ensuring everyone is equipped with the latest capabilities. - Educational Institutions that aim to provide more real-world, hands-on experience for their students. To fellow B2B tech founders: always focus on your customers. Make them your friends, understand their pain points, and build products around their needs. Your traction and customer stories will drive your journey forward. This is just the beginning for LabLabee! 🌍 More details here : https://lnkd.in/d5Nh3SBY #StartupLife #EdTech #Telecom #Funding

382

64 Comments

#GenAI has been so buzzy the past year, it's hard to sort out the wheat from the chaff (not that I've been doing that manually lately...robotic process automation with GenAI helps). In any case, take a quick read of my most recent article below on #telcos and (what else) #genAI and LMK what you think. P.S. You can come up to me at FutureNet World in London in the next two days (16-17 Apr) and tell me what you *really* think 😀

9

2 Comments

Over the past few years, I’ve often been asked, “Are you an AI company?” My response has always been a clear “No.” We are a core product company that leverages AI where it makes sense. While we’ve explored AI investments, we’ve stayed true to our foundation—building a strong core product with a focus on profitability in accelerating digital infrastructure deployment. Fast forward to 2024, and this approach has truly paid off. As AI models become rapidly commoditized, our data has become our competitive moat! Now, we’re able to fine-tune top-tier models to deliver exceptional value to our customers and investors at an accelerated pace. 💪 I am sure other product companies have had similar decision points - sticking to fundamentals is the best way to build a business! I decided to write about my reasonings in this article (4 min read): https://lnkd.in/gAc6z7EU #AI #DigitalTransformation #DataDriven #CoreProduct #TechInnovation #AI-Enabled

96

5 Comments