T-Mobile US CTO John Saw revealed it is testing whether humanoid robots could respond faster and rely less on onboard computers by running some of their AI at its network edge.

In a blog, Saw explained T-Mobile is working with robotics company Figure AI to test if its systems can run on nearby computing infrastructure without slowing the machines’ responsiveness.

The systems involve controlling robots’ movement, and interpreting visual information and language instructions.

Saw wrote shifting some processing off the robots could reduce battery use, computing demands, heat, weight and cost. It could also let them run more capable AI without needing more hardware or power.

AWS
T-Mobile is working with AWS on the first deployments of Kinetic Edge, the operator’s distributed computing infrastructure for physical AI.

They are using T-Mobile’s 5G-Advanced network and distributed edge along with AWS’ cloud orchestration and AI services to push “deterministic connectivity, compute and intelligence closer to where machines operate”, Saw wrote.

He explained reliable connections and localised computing could help businesses take physical AI beyond individual demonstrations, highlighting developers are putting vision and reasoning models at the network edge to make robots, cameras and sensors less dependent on a distant cloud.

“This early work is helping us explore the capabilities that could make physical AI possible at scale, while opening the door to entirely new applications and opportunities.”

Machines require consistent connectivity when moving, sufficient uplink capacity to send sensor data, low latency and wide-area coverage, Saw wrote. Wi-Fi and satellite have a role, but he argued advanced terrestrial wireless networks are better suited to meeting the needs.

Human interaction
Saw noted employing people to assist robots would become challenging when thousands of machines are involved: T-Mobile is working with Serve Robotics to evaluate the feasibility of conducting monitoring at the edge to assess which problems can be solved by humans and which need more AI input.

The executive also highlighted work with Nvidia, Nokia and a growing group of developers to run physical AI applications across distributed edge networks, along with the recent completion of an AI-RAN call involving the vendors.

Saw explained T-Mobile is working with the city of San Jose in the state of California and specialist companies to develop a way to create an overview of traffic using edge AI to combine the views of separate 5G cameras.

They are testing if AI can connect occurrences at different junctions and if providing a big-picture could help people spot traffic patterns sooner.

The operator is also exploring “how AI and network workloads can share accelerated computing infrastructure” through partnerships with Nvidia, Nokia and Ericsson.

Saw said T-Mobile must develop ways to deliver “data and workloads” to “where intelligence needs to run”, whether it be on devices, the edge or cloud.

He said the direction “is becoming clearer” and T-Mobile’s network would “be the connective tissue” between the layers.

The post T-Mobile US pushes physical AI to the edge appeared first on Mobile World Live.

The prime minister of Singapore backed the city state to carve out a meaningful role in industrial AI, arguing it holds several advantages over frontier developments being led by companies in the US and China.

Prime Minister Lawrence Wong (pictured, right) said Singapore holds several advantages in applying AI to specific sectors compared with developers in other nations working on more general uses.

“We are small. We are compact. We are highly connected as a society”, Prime Minister Wong said as he detailed a collaboration with Germany to develop industrial AI uses.

The prime minister argued Singapore’s clear laws and “protection for IP” also stand it in good stead.

“We see a lot of opportunities in sectors like healthcare, logistics, finance and advanced manufacturing”.

German Chancellor Friedrich Merz (pictured, left) and Prime Minister Wong put industrial AI development at the heart of future collaboration goals.

The Singaporean leader said the nations are on common ground over protecting from the risks AI poses, pledging development would be “done in a way that remains safe, trusted and human centred”.

Advanced manufacturing, sustainable technologies, semiconductors, and quantum and space technologies are among the other fields the nations intend to collaborate on.

Prime Minister Wong predicted joint development of industrial AI would result in “more robotics and automated solutions in manufacturing and logistics”, and have impacts on “other areas of the economy”.

He noted trade agreements with Germany are positive in a global context, noting long-standing relationships between companies in each nation and backing Europe as a whole as a solid long-term investment bet.

The post Singapore, Germany target applied AI appeared first on Mobile World Live.

Anthropic filed for an IPO which could rival SpaceX’s record-breaking listing, as the AI business revealed a rise in revenue and billions of dollars of infrastructure cost commitments.

The Claude large language model (LLM) developer’s S-1 filing with the US Securities and Exchange Commission yesterday (26 October) offered the first detailed public look at its finances. Anthropic submitted IPO paperwork confidentially in June, allowing the agency to review the figures before their wider release.

Anthropic applied to list Class A shares on the Nasdaq Global Select Market under the symbol ANT but did not disclose how many shares it would offer, a price range or valuation. Goldman Sachs and Morgan Stanley headed the list of underwriters.

The documents revealed revenue of $16.3 billion in the first half of 2026, up from $1.2 billion in H1 2025 and well above the $4.6 billion recorded for the whole of 2025.

Cost of revenue rose to $2.4 billion from just over $1 billion, with compute and infrastructure accounting for almost all of it. Anthropic cautioned operating margins will fluctuate as more capacity comes online.

The company booked a net loss of $80.7 billion in H1. Of the amount, $72.4 billion reflected non-cash charges from revaluing convertible notes and similar instruments as its valuation rose.

Amazon and Google supply key parts of Anthropic’s computing infrastructure.

It runs workloads on Amazon Web Services (AWS) Trainium chips, Google Tensor Processing Units, and GPUs from Nvidia and AMD. It also pays fees to sell Claude through AWS, Google Cloud and Microsoft Azure.

Amazon holds convertible notes which will convert into non-voting Class N shares at the IPO, subject to an ownership cap.

Anthropic’s infrastructure plans include spending at least $110 billion on Amazon cloud services until April 2036 and up to $120.5 billion with SpaceX through 2029.

It has committed a further $114.5 billion to Google until 2033 and $44.9 billion to Microsoft until May 2033 for computing infrastructure.

Risks
In the documents Anthropic warned some cloud suppliers are developing rival AI models and could restrict, reprice or cut off its access to computing capacity.

It also highlighted criminals have already used AI in cyberattacks and warned misuse in biology could cause greater harm.

The company also highlighted the potential for AI to concentrate power and destabilise geopolitics.

Anthropic increasingly uses its models to develop their successors and noted the new versions’ capabilities could advance faster than safety measures. It also cautioned AI systems could act independently in harmful ways if they fail to follow intended goals.

Rival OpenAI confidentially filed paperwork for a US IPO in June, but delayed the listing until early 2027.

The post Anthropic files for IPO, reveals revenue surge appeared first on Mobile World Live.

• Infrastructure startup Hyperbolic raised a $30 million round led by Intel CEO Lip-Bu Tan's VC firm.
• The startup connects companies with neoclouds and makes their AI chips easier to use.
• Hyperbolic also wants to let customers resell unused computing power.

Intel CEO Lip-Bu Tan is backing an infrastructure startup that has raised $30 million to help companies rent compute across hundreds of different AI cloud providers.

Hyperbolic, an AI chip marketplace and software startup, has raised $50 million to date.

Walden Catalyst, the venture capital firm founded by Tan and former Samsung president Young Sohn, led the Series A-1 funding round. Dell Technologies Capital, General Catalyst, Vista Equity Partners, and Cambium Capital also participated.

As companies scramble for computing power, one challenge is that AI cloud providers, or neoclouds, use a mix of different chips, networking, and software. Hyperbolic helps customers find compute resources from different providers and creates software that makes it easier to use them all, rather than dealing with each one separately.

"It's been a really fragmented market that makes the customer experience really painful," said 31-year-old founder and CEO Jasper Zhang.

Hyperbolic also wants to let customers resell compute that they're paying for but no longer need.

The four-year-old startup works closely with about 30 neoclouds, including WhiteFiber and Vector Core Compute. It says it has roughly $200 million worth of compute connected to its software annually.

Hyperbolic previously raised a $12 million Series A round in 2024, led by Variant and Polychain Capital, following a $7 million seed round earlier that year. The company expects to have 50 employees by the end of the year and will use its funding to expand software tools for managing AI chips across different neoclouds.

Zhang said the company is aiming for $150 million in annual recurring revenue by the end of the year and is already breaking even. More than 250,000 developers, including teams at Modal, Higgsfield, and Daytona, use Hyperbolic.

Before founding the company, Zhang worked as a quantitative researcher at Citadel Securities. As Wall Street explores ways to financialize AI chips, he said his startup's plans to let companies resell unused compute could help pave the way for a futures market.

Here's a look at the pitch deck Hyperbolic used to raise its $30 million Series A-1. Slides have been redacted so that the deck can be shared publicly.