Astor Perkins

Quantum networks with coherent routing of information through multiple nodes Large-scale communication networks, such as the Internet, rely on routing packets of data through multiple intermediate nodes to transmit information from a sender to a receiver. In this paper, we develop a model of a quantum communication network that routes information simultaneously along multiple paths passing through intermediate stations. We demonstrate that a quantum routing approach can in principle extend the distance over which information can be transmitted reliably. Surprisingly, the benefit of quantum routing also applies to the transmission of classical information: even if the transmitted data is purely classical, delocalising it on multiple routes can enhance the achievable transmission distance. Our findings highlight the potential of a future quantum internet not only for achieving secure quantum communication and distributed quantum computing but also for extending the range of classical data transmission. https://lnkd.in/esHZkbyh

Nature: How China created AI model DeepSeek and shocked the world Earlier this week, DeepSeek launched another model, called Janus-Pro-7B, which can generate images from text prompts much like OpenAI’s DALL-E 3 and Stable Diffusion, made by Stability AI in London. If DeepSeek-R1’s performance surprised many people outside of China, researchers inside the country say the start-up’s success is to be expected and fits with the government’s ambition to be a global leader in AI. It was inevitable that a company such as DeepSeek would emerge in China, given the huge venture-capital investment in firms developing LLMs and the many people who hold doctorates in science, technology, engineering or mathematics fields, including AI, says Yunji Chen, a computer scientist working on AI chips at the Institute of Computing Technology of the Chinese Academy of Sciences in Beijing. “If there was no DeepSeek, there would be some other Chinese LLM that could do great things.” In fact, there are. On 29 January, tech behemoth Alibaba released its most advanced LLM so far, Qwen2.5-Max, which the company says outperforms DeepSeek's V3, another LLM that the firm released in December. And last week, Moonshot AI and ByteDance released new reasoning models, Kimi 1.5 and 1.5-pro, which the companies claim can outperform o1 on some benchmark tests. In 2017, the Chinese government announced its intention for the country to become the world leader in AI by 2030. It tasked the industry with completing major AI breakthroughs “such that technologies and applications achieve a world-leading level” by 2025. Developing a pipeline of ‘AI talent’ became a priority. By 2022, the Chinese ministry of education had approved 440 universities to offer undergraduate degrees specializing in AI, according to a report from the Center for Security and Emerging Technology (CSET) at Georgetown University in Washington DC. In that year, China supplied almost half of the world’s leading AI researchers, while the United States accounted for just 18%, according to the think tank MacroPolo in Chicago. DeepSeek probably benefited from the government’s investment in AI education and talent development, which includes numerous scholarships, research grants and partnerships between academia and industry. For instance, state-backed initiatives such as the National Engineering Laboratory for Deep Learning Technology and Application, which is led by tech company Baidu, have trained thousands of AI specialists. Exact figures on DeepSeek’s workforce are hard to find, but company founder Liang Wenfeng told media that the company has recruited graduates and doctoral students from top-ranking Chinese universities. “They are deeply motivated by a drive for self-reliance in innovation.” He co-founded the hedge fund High-Flyer almost a decade ago and established DeepSeek in 2023. https://lnkd.in/eqByfM23

Japan's Resilience moon lander views Earth's most remote region from orbit The second lunar lander from Japanese space exploration company ispace snapped a stunning image of the Earth as it prepares for its journey to the moon. ispace's Resilience lander launched on Jan. 15 atop a SpaceX Falcon 9 rocket, sharing the ride with Firefly Aerospace's Blue Ghost lander. Resilience is still in orbit around the Earth and is gradually raising its orbit ready to shoot for the moon. The lander's energy-efficient path will see Resilience attempt to set down in Mare Frigoris ("Sea of Cold") in the northern hemisphere of the moon in around four months from now. In an update on Jan. 29, ispace said Resilience is in excellent health and continuing its journey in Earth orbit. What's more, the spacecraft has returned a stark image of our home planet from a unique perspective, showing the most remote area of the Earth. "Resilience knows what it means to be alone in the vastness of space," an ispace post on social media X stated. "Looking back at Earth on Jan. 25, 2025, the lander was about 10,000km [6,000 miles] from our Blue Marble, poignantly capturing Point Nemo, the most remote place on our planet, about 2,688 kilometers [1,670 miles] from the nearest land." Resilience is based on ispace's Hakuto-R solar-powered lander platform. It is similar to the first lander which made a failed landing attempt in April 2023, but features upgraded software to overcome sensor issues. It carries a range of commercial and science payloads, as well as a microrover, named Tenacious. The compact 11 pounds (5 kilograms) mobile spacecraft will collect samples with a small shovel and analyze these with an imaging camera. While Resilience and Blue Ghost are in orbit and preparing to leave for the moon, another lunar lander — Intuitive Machines' IM-2, or Athena — recently arrived at Cape Canaveral to loaded atop of a SpaceX Falcon 9 rocket, ready for a launch window opening no earlier than Feb. 26. Resilience itself will also take onboard lessons from ispace's first lunar landing attempt, which crashed into the moon in April 2023, due to an onboard altitude sensor being confused by the rim of a crater The company is also building a larger lander, named Apex 1.0. Its first outing will be ispace's Mission 3, expected to launch around 2026. https://lnkd.in/eCtYqJZa

Why ‘Distillation’ Has Become the Scariest Word for AI Companies DeepSeek’s success learning from bigger AI models raises questions about the billions being spent on the most advanced technology Tech giants have spent billions of dollars on the premise that bigger is better in artificial intelligence. DeepSeek’s breakthrough shows smaller can be just as good. The Chinese company’s leap into the top ranks of AI makers has sparked heated discussions in Silicon Valley around a process DeepSeek used known as distillation, in which a new system learns from an existing one by asking it hundreds of thousands of questions and analyzing the answers. https://lnkd.in/eH46gt4e

This quantum computer built on server racks paves the way to bigger machines A Canadian startup called Xanadu has built a new quantum computer it says can be easily scaled up to achieve the computational power needed to tackle scientific challenges ranging from drug discovery to more energy-efficient machine learning. Aurora is a “photonic” quantum computer, which means it crunches numbers using photonic qubits—information encoded in light. In practice, this means combining and recombining laser beams on multiple chips using lenses, fibers, and other optics according to an algorithm. Xanadu’s computer is designed in such a way that the answer to an algorithm it executes corresponds to the final number of photons in each laser beam. This approach differs from one used by Google and IBM, which involves encoding information in properties of superconducting circuits. Aurora has a modular design that consists of four similar units. Ultimately, Xanadu envisions a quantum computer as a specialized data center, consisting of rows upon rows of these servers. This contrasts with the industry’s earlier conception of a specialized chip within a supercomputer, much like a GPU. But this work, which the company published last week in Nature, is just a first step toward that vision. Aurora used 35 chips to construct a total of 12 quantum bits, or qubits. Any useful applications of quantum computing proposed to date will require at least thousands of qubits, or possibly a million. By comparison, Google’s quantum computer Willow, which debuted last year, has 105 qubits, and IBM’s Condor has 1,121. Xanadu’s 12 qubits may seem like a paltry number next to IBM’s 1,121, but Tiwari says this doesn’t mean that quantum computers based on photonics are running behind. Photonic quantum computers offer several design advantages. The qubits are less sensitive to environmental noise, says Tiwari, which makes it easier to get them to retain information for longer. It is also relatively straightforward to connect photonic quantum computers via conventional fiber optics, because they already use light to encode information. Networking quantum computers together is key to the industry’s vision of a “quantum internet” where different quantum devices talk to each other. Aurora’s servers also don’t need to be kept as cool as superconducting quantum computers, says Weedbrook, so they don’t require as much cryogenic technology. The server racks operate at room temperature, although photon-counting detectors still need to be cryogenically cooled in another room. Xanadu is not the only company pursuing photonic quantum computers; others include PsiQuantum in the US and Quandela in France. Other groups are using materials like neutral atoms and ions to construct their quantum systems. https://lnkd.in/g467ftCT

Europe awards $900 million contract for Argonaut lunar lander development Thales Alenia Space announced a contract worth nearly $900 million to develop and deliver the Lunar Descent Element (LDE) for Argonaut, the European Space Agency’s cargo lander slated for missions to the Moon starting in the 2030s. The contract includes mission design and integration of the LDE, which would be responsible for transporting and landing the spacecraft on the Moon. The LDE would be joined by an adaptable interface element designed to support a wide variety of cargo and scientific payloads. Thales Alenia Space, a joint venture between France’s Thales and Italy’s Leonardo, said Argonaut’s first mission is expected to deliver navigation and telecommunication payloads, along with an energy generation and storage system, supporting commercial European exploration of the lunar south pole The European Space Agency is also positioning Argonaut, which has a capacity of approximately two metric tons of cargo, as a potential asset for future NASA Artemis lunar missions. “This new element of the Artemis program will facilitate long-duration human lunar exploration missions and be crucial for enhancing European autonomy in lunar exploration,” said Thales Alenia Space CEO Hervé Derrey. Thales Alenia Space emphasized Argonaut’s versatility, with an interface designed to support a wide range of missions. According to the company, the lander could carry supplies for astronauts, deploy rovers, host technology demonstrations, support lunar resource utilization, and accommodate a telescope or power station. While Thales Alenia Space is the prime contractor for LDE development, overall mission responsibility, including payload integration and LDE operations, will be determined through a separate procurement process in the future. https://lnkd.in/g7T4Qq6H

Cerebras becomes the world’s fastest host for DeepSeek R1, outpacing Nvidia GPUs by 57x Cerebras Systems announced today it will host DeepSeek’s breakthrough R1 artificial intelligence model on U.S. servers, promising speeds up to 57 times faster than GPU-based solutions while keeping sensitive data within American borders. The move comes amid growing concerns about China’s rapid AI advancement and data privacy. The AI chip startup will deploy a 70-billion-parameter version of DeepSeek-R1 running on its proprietary wafer-scale hardware, delivering 1,600 tokens per second — a dramatic improvement over traditional GPU implementations that have struggled with newer “reasoning” AI models. Cerebras achieves its speed advantage through a novel chip architecture that keeps entire AI models on a single wafer-sized processor, eliminating the memory bottlenecks that plague GPU-based systems. The company claims its implementation of DeepSeek-R1 matches or exceeds the performance of OpenAI’s proprietary models, while running entirely on U.S. soil. The development represents a significant shift in the AI landscape. DeepSeek, founded by former hedge fund executive Liang Wenfeng, shocked the industry by achieving sophisticated AI reasoning capabilities reportedly at just 1% of the cost of U.S. competitors. Cerebras’ hosting solution now offers American companies a way to leverage these advances while maintaining data control. “It’s actually a nice story that the U.S. research labs gave this gift to the world. The Chinese took it and improved it, but it has limitations because it runs in China, has some censorship problems, and now we’re taking it back and running it on U.S. data centers, without censorship, without data retention,” Wang said. The service will be available through a developer preview starting today. While it will be initially free, Cerebras plans to implement API access controls due to strong early demand. Industry analysts suggest this development could accelerate the shift away from GPU-dependent AI infrastructure. “Nvidia is no longer the leader in inference performance,” Wang noted, pointing to benchmarks showing superior performance from various specialized AI chips. “These other AI chip companies are really faster than GPUs for running these latest models.” The impact extends beyond technical metrics. As AI models increasingly incorporate sophisticated reasoning capabilities, their computational demands have skyrocketed. Cerebras argues its architecture is better suited for these emerging workloads, potentially reshaping the competitive landscape in enterprise AI deployment. https://lnkd.in/gw6xE48V

What We Learned About the Future of AI from Microsoft and Meta Earnings Key Takeaways - Microsoft, Meta, and IBM all highlighted healthy growth in their AI businesses in the last three months of 2024. - Microsoft and Meta, which cumulatively expect to invest nearly $150 billion in infrastructure in their respective 2025 fiscal years, stood by their aggressive spending plans. - Executives expressed confidence that Chinese start-up DeepSeek's incredibly cost-efficient AI model, which spooked investors earlier this week, was ultimately a good thing for their businesses. https://lnkd.in/gHA43gEx

FDA Approves Journavx Drug to Treat Pain Without Addiction Risk The drug, Journavx by Vertex Pharmaceuticals, blocks pain signals to the brain, making it nonaddictive. https://lnkd.in/gCGqicb2

Integrating single-cell RNA and T cell/B cell receptor sequencing with mass cytometry reveals dynamic trajectories of human peripheral immune cells from birth to old age A comprehensive understanding of the evolution of the immune landscape in humans across the entire lifespan at single-cell transcriptional and protein levels, during development, maturation and senescence is currently lacking. We recruited a total of 220 healthy volunteers from the Shanghai Pudong Cohort (NCT05206643), spanning 13 age groups from 0 to over 90 years, and profiled their peripheral immune cells through single-cell RNA-sequencing coupled with single T cell and B cell receptor sequencing, high-throughput mass cytometry, bulk RNA-sequencing and flow cytometry validation experiments. We revealed that T cells were the most strongly affected by age and experienced the most intensive rewiring in cell–cell interactions during specific age. Different T cell subsets displayed different aging patterns in both transcriptomes and immune repertoires; examples included GNLY+CD8+ effector memory T cells, which exhibited the highest clonal expansion among all T cell subsets and displayed distinct functional signatures in children and the elderly; and CD8+ MAIT cells, which reached their peaks of relative abundance, clonal diversity and antibacterial capability in adolescents and then gradually tapered off. Interestingly, we identified and experimentally verified a previously unrecognized ‘cytotoxic’ B cell subset that was enriched in children. Finally, an immune age prediction model was developed based on lifecycle-wide single-cell data that can evaluate the immune status of healthy individuals and identify those with disturbed immune functions. Our work provides both valuable insights and resources for further understanding the aging of the immune system across the whole human lifespan. https://lnkd.in/eRQqKSeC