Today on R&D World
Anthropic says Fable 5.2 nearly doubles its predecessor’s hardware-design score. Token prices fell 20%, but the cost per task doubled at full effort.
Carnyx targets 2027 IND filing for oral sleep candidate CNYX-005
OpenAI unveils BioBasicsBench, a benchmark to gauge how well AI agents read raw experimental data across 100 routine wet-lab tasks.
Edison’s Kosmos AI scientist moves into Pfizer R&D; Edison details the math behind its $99 million-per-program estimate
Isomorphic Labs president says all drugs will be AI-designed as first trials near a year-end target
10x Genomics CTO expects AI to leave some of biomedical R&D ‘painfully slow’ while other parts are ‘massively accelerated’
NIH proposes one safety-monitoring standard for all its clinical research
Sam Altman says a machine for scientific discovery is closer than he expected
Inside the Lab: Can AI Solve the Critical Materials Challenge?
Two Science papers catch altermagnets by handedness
10x Genomics ships Sentira Single Cell, a Claude agent for the ‘last mile’ of single-cell analysis
Harvard-led team says NIH-funded healthy-tissue atlas is AI-ready, but its paper warns it’s not fully representative
DOE’s fusion voucher program opens to consortia and pilot-plant reviews
Agilent completes $240 million Proscia acquisition, taking ownership of the pathology software it distributed
Physics See More >
Two Science papers catch altermagnets by handedness
Manganese fluoride (MnF₂) was one of the first materials in which scientists used neutrons to observe antiferromagnetic order, a pattern where neighboring atoms’ spins lock into an alternating up-down pattern, and neutron-scattering courses have used it as a textbook antiferromagnet. Now, it is one of two materials that new papers in the journal Science identify…
LLNL’s gamma-ray sensor is bound for a Martian moon
OpenAI says its AI models produced major proofs in mathematics, computer science and physics
Northwestern and Western Digital demonstrate a new way to write magnetic memory
How a Duke-led HPC method brought cell-scale cancer simulation to a single cloud node
Sponsored Content See More >
Operational discipline is a critical determinant of clinical utility for neurologic monoclonal antibodies
By Bill Holt, DO, Vice President, Neuroscience Therapeutic Area, ICON For neurologic monoclonal antibodies (mAbs), the scientific hurdle is no longer the primary barrier to success. Increasingly, operational execution—not efficacy—determines whether these therapies reach patients at scale. From multiple sclerosis (MS) to neuromyelitis optica spectrum disorder (NMOSD) to early Alzheimer’s disease, mAbs have moved from…
Life Science See More >
Google is teaching healthcare professionals to incorporate AI into the everyday practice of medicine
Google’s four-course AI for Healthcare Professionals Certificate covers patient communication, admin work and vibe coding, with checks on AI output.
Carnyx targets 2027 IND filing for oral sleep candidate CNYX-005
OpenAI unveils BioBasicsBench, a benchmark to gauge how well AI agents read raw experimental data across 100 routine wet-lab tasks.
Edison’s Kosmos AI scientist moves into Pfizer R&D; Edison details the math behind its $99 million-per-program estimate
10x Genomics CTO expects AI to leave some of biomedical R&D ‘painfully slow’ while other parts are ‘massively accelerated’
Nanotechnology See More >
Researchers developed quantum nanosensors that can measure the temperature of a single cell
Researchers have developed molecular quantum nanosensors (MoQNs) designed to operate in the cytoplasm and nuclei of living cancer cells to map radical-generation processes and thermal dynamics that are linked to cancer-associated cellular physiology. The sensors use molecular-level uniformity to achieve a threefold enhancement in spectral resolution and superior thermometric specificity. The platform enables absolute temperature…
Energy See More >
Inside the Lab: Can AI Solve the Critical Materials Challenge?
The technologies shaping the future—from electric vehicles and wind turbines to fusion energy, robotics, and national defense—depend on critical minerals and advanced materials. Yet discovering and deploying new materials can take years or even decades, while technological progress and demand accelerate. This challenge is compounded by supply chains vulnerable to geopolitical pressures, resource constraints, and…
DOE’s fusion voucher program opens to consortia and pilot-plant reviews
DOE’s BattMan brings four national labs together to de-risk battery scale-up
INL’s $60 million nuclear AI project will spend the first year testing where AI agents can help and where they might hurt
One SwRI R&D 100 winner turns buckled fiberglass into a spacecraft truss. The other models freight rail with 900 hours of BNSF field data
Chemistry See More >
CAS AI leader on bringing agents into scientific workflows, with a Microsoft assist
CAS AI leader Andrea Jacobs on bringing CAS data and its Newton agent into Microsoft Copilot through MCP, and on connecting AI agents to lab instruments.
150 years of CAS chemistry data comes to Microsoft Copilot
AI is wreaking havoc on math and coding. Where could it have the biggest impact next?
R&D 100 Researcher of the Year Parans Paranthaman on superconducting tape, oilfield lithium and getting discoveries into production
PNNL scientists used ultrafast X-ray spectroscopy to capture PCET reaction in new detail
Material Science See More >
Inside the Lab: Can AI Solve the Critical Materials Challenge?
The technologies shaping the future—from electric vehicles and wind turbines to fusion energy, robotics, and national defense—depend on critical minerals and advanced materials. Yet discovering and deploying new materials can take years or even decades, while technological progress and demand accelerate. This challenge is compounded by supply chains vulnerable to geopolitical pressures, resource constraints, and…
Two Science papers catch altermagnets by handedness
Northwestern and Western Digital demonstrate a new way to write magnetic memory
Eggshell-inspired shield cut projectile speed by 65% in simulation, but light-gas gun testing has not confirmed it
One SwRI R&D 100 winner turns buckled fiberglass into a spacecraft truss. The other models freight rail with 900 hours of BNSF field data
Semiconductors See More >
OpenAI says GPT-6 Astra designed a talking Codex “pet.” Manufacturing took several rounds
OpenAI says GPT-6 Astra designed the case, circuit board and firmware for a talking Codex device. A factory built prototypes after several rounds of fixes.
Northwestern and Western Digital demonstrate a new way to write magnetic memory
Google teases Synaptics edge AI chip ahead of Oct. 20 launch
Materials startup Thintronics takes aim at Ajinomoto’s chip packaging film, with surging AI demand widening the opening
OpenAI and Synopsys plan a specialized model for chip design
Aerospace See More >
LLNL’s gamma-ray sensor is bound for a Martian moon
An LLNL germanium sensor aboard JAXA’s MMX spacecraft will map the elements on Phobos and help pick where the mission collects samples for return in 2031.
The UAE’s new space probe will crash into the asteroid it’s designed to study
Semiconductors, biomanufacturing and dual-use tech fill ISS National Lab’s startup cohort
Eggshell-inspired shield cut projectile speed by 65% in simulation, but light-gas gun testing has not confirmed it
