Financial Times | Energy & Natural Resources
“Japan’s TDK is claiming a breakthrough in materials used in its small solid-state batteries, with the Apple supplier predicting significant performance increases for devices from wireless headphones to smartwatches.
The new material provides an energy density — the amount that can be squeezed into a given space — of 1,000 watt-hours per litre, which is about 100 times greater than TDK’s current battery in mass production.”
From Financial Times.
“A Chinese-built large language model called DeepSeek-R1 is thrilling scientists as an affordable and open rival to ‘reasoning’ models such as OpenAI’s o1.
These models generate responses step-by-step, in a process analogous to human reasoning. This makes them more adept than earlier language models at solving scientific problems and could make them useful in research. Initial tests of R1, released on 20 January, show that its performance on certain tasks in chemistry, mathematics and coding is on par with that of o1 — which wowed researchers when it was released by OpenAI in September…
DeepSeek hasn’t released the full cost of training R1, but it is charging people using its interface around one-thirtieth of what o1 costs to run. The firm has also created mini ‘distilled’ versions of R1 to allow researchers with limited computing power to play with the model. An ‘experiment that cost more than £300 with o1, cost less than $10 with R1,’ says Krenn.”
From Nature.
“Researchers have designed and built the world’s first microscope made entirely from 3D-printed parts. And because the open-source plans are already available online, almost anyone can assemble their own for barely $60.
A few years ago, researchers at the University of Bath and University of Cambridge started the OpenFlexure project offering open-source blueprints for 3D-printed microscopes. With its core development group now based at the University of Glasgow, OpenFlexure microscopes have been assembled in over 50 countries around the world, as well as at laboratories in Antarctica. While much cheaper than standard equipment, there was a caveat: a microscope’s specially crafted glass lenses often cost hundreds of dollars, putting the tools out of many people’s price range. Recently, however, a team at the University of Strathclyde developed a workaround they say both lowers the OpenFlexure design’s total cost, and makes it the first completely 3D-printed microscope.”
From 3DPrint.com.
“Artificial intelligence has advanced rapidly over the past 15 years, fueled by the success of deep learning.
A key reason for the success of deep learning systems has been their ability to keep improving with a staggering increase in the inputs used to train them — especially computation.
Before deep learning took off around 2010, the amount of computation used to train notable AI systems doubled about every 21 months. But, as you can see in the chart, this has accelerated significantly with the rise of deep learning, now doubling roughly every six months.
As one example of this pace, compared to AlexNet, the system that represented a breakthrough in computer vision in 2012, Google’s system ‘Gemini 1.0 Ultra’ just 11 years later used 100 million times more training computation.”
From Our World in Data.
“Drugs developed by Alphabet’s drug discovery subsidiary and designed by artificial intelligence are expected to head to trial by the end of the year, according to a Google executive…
The four-year-old Isomorphic Labs was spun off from DeepMind in 2021 as a stand-alone subsidiary under Alphabet. In July, it announced deals to work on research with Eli Lilly & Co. and Novartis to leverage its AI technology — namely AlphaFold, its model that predicts a protein’s 3D structure — to discover therapeutics against multiple targets.
‘We’re looking at oncology, cardiovascular, neurodegeneration, all the big disease areas, and I think by the end of this year, we’ll have our first drug,’ Hassabis told The Financial Times on Tuesday.”
From Quartz.
