What’s up with… Amazon & Anthropic, Odido, Verizon & Ciena

• Amazon doubles its Anthropic investment, takeover looks likely 
• Dutch operator Odido is heading for an IPO
• Ciena helps Verizon deal with AI workloads

In today’s industry news roundup: Amazon has upped its investment in GenAI specialist Anthropic to $8bn, a move that signals an eventual takeover, reckons analyst; Odido’s private equity owners are lining up an IPO for the Netherlands telco; Verizon hits 1.6 Tbit/s over a single wavelength using Ciena gear; and much more!

Amazon has doubled its investment in generative AI (GenAI) developer Anthropic to $8bn and strengthened its ties with the company in a move that looks like a precursor to a takeover. Earlier this year, Amazon completed its initial $4bn investment in Anthropic: Now it has pumped another $4bn into the company and expanded the relationship. “Anthropic is now naming AWS [Amazon Web Services] its primary training partner, in addition to continuing to be its primary cloud provider, and will use AWS Trainium and Inferentia chips to train and deploy its future foundation models,” Amazon announced in this blog. “Both companies will continue to work closely to keep advancing Trainium’s hardware and software capabilities,” it added. The move clearly signals that Anthropic will eventually be swallowed up by the big tech firm, reckons seasoned industry analyst Richard Windsor. “Amazon has upped its investment in Anthropic, which virtually guarantees that when the money runs out, Anthropic will be acquired and become the AI department of Amazon, creating a fully vertically integrated player ready to fight for the AI ecosystem prize… Amazon has added another $4bn to its investment in Anthropic, bringing the total to $8bn, which cements the virtual certainty that Amazon will end up acquiring Anthropic,” he stated in his latest Radio Free Mobile blog, adding that the battle lines are being drawn between Amazon and Microsoft (with its OpenAI investment), Meta, Google and Apple. 

Dutch operator Odido, which used to be known as the more staid-sounding T-Mobile Netherlands under Deutsche Telekom ownership (even after its merger with local rival Tele2) before an eventual rebranding by its new owners Warburg Pincus and Apax Partners, is being lined up for an IPO that could value it at €7bn, according to Bloomberg (as reported by Reuters). 

US telco Verizon achieved data transport speeds of up to 1.6 Tbit/s over single optical wavelength using Ciena’s WaveLogic 6 Extreme coherent optical solution during a recent trial over its live network in Boston. Verizon cited the networking breakthrough as important in an era when AI workloads are putting extra strain on communications networks. “AI is contingent on analysing billions of data points in real time on an ongoing basis. Because of the massive, multi-year transformation on Verizon’s network, including deploying cloud-native architecture from the core to the edge, deploying massive amounts of high-performing spectrum, exponentially increasing the capacity of our fibre infrastructure and deploying advanced technologies and intelligence throughout the network, Verizon’s network provides the ability to power the processes and movements of AI-generated activity,” said Adam Koeppe, SVP of technology strategy and planning at Verizon. “This continued advancement of our fibre network will further position us to be the provider of choice for AI workloads now and in the future.” Read more. 

The Telecom Regulatory Authority of India (TRAI) has directed all the sub-continent’s mobile service operators to publish geospatial maps on their websites to provide the “much needed” transparency required to empower subscribers to decide which company in their particular geographic locality to opt for when signing up to a service provider. The new ruling is part of the TRAI’s push to provide consumers with clear and detailed information about the availability of all telecom services across India. The new maps will show the geographical areas covered by all mobile service providers, including the particular generation of mobile technology available, be that 2G, 3G, 4G and/or 5G. By mandating standardised geographical service coverage be made available to the public, the TRAI directive places a new responsibility and accountability on India’s mobile operators which, the regulator expects, will actually further competition across the massive industry. The new regulation is central to the TRAI’s emphasis on ensuring mobile operators provide the best possible quality of service (QoS) to their subscribers. To ensure compliance, the constantly updated geospatial maps must show customers the percentage of working cell availability. The benchmark is 99%. Furthermore, the information displayed on the maps must be real and tabulated from either actual measurement of each cell site or the physical parameters of network analytics. India’s Financial Express reports that the TRAI regulation also requires “the cell coverage of respective technologies (2G/3G/4G/5G) [to] be presented on the coverage map in the prescribed colour scheme, having the minimum prescribed signal strength” while a “link for the coverage map (with suitable logo) be provided on the home or landing page of the service provider’s website at a prominent location for single-click navigation to ensure adequate visibility and ease of access.”

The ex-CEO of Google, Eric Schmidt, recently paid a return visit to his ivy-league alma mater, Princeton University, in New Jersey in the US to speak to an audience of undergraduate, postgraduate students plus assorted academic worthies and to promote the book “Genesis: Artificial Intelligence, Hope and the Human Spirit”, which he co-authored with Craig Mundle, the former CTO of Microsoft, and the late Henry Kissinger. He told the gathering  that the great majority of people will be unable to understand, fully appreciate or cope with the technological and societal changes that AI will engender. “I can assure you that the humans in the rest of the world, all the normal people – because you all are not normal, sorry to say, you’re special in some way – the normal people are not ready. Their governments are not ready. The government processes are not ready. The doctrines are not ready. They’re not ready for the arrival of this,” he said. As reported in the Princeton Alumni Weekly, Schmidt thinks it more than likely that in the fairly near future, AI will shape a child’s identity, cultural norms and worldview via the persona of what would be a non-human (ie. robotic) AI “best friend”. Presumably, there’ll soon no longer be any need for a child at a certain stage of its development to have imaginary best friends with whom to interact and play and learn as they have done for more than half a million years. He asked his audience the rhetorical question, “So, what are the rules? Is it OK that it’s the equivalent of Mark Zuckerberg as the surrogate parent who gets to decide what your kid learns and doesn’t learn?” The resounding, evident and unrhetorical answer must surely be, “No it ***** well isn’t!” Schmidt stressed the pressing importance of globally agreed safety parameters for AI, commenting, “Playing with the way people think is really powerful. If you think about state-sponsored misinformation, that’s trivial compared to having your best friend be state-sponsored, and they sort of have daily interaction and shape someone’s identity, their cultural values.” Wonder where he was talking about?

– The staff, TelecomTV