Tech News Review Week 9-14 January 2021, by Stelios Iordanis

Energy

New nanostructured alloy for anode is a big step toward revolutionizing energy storage

https://phys.org/news/2021-01-nanostructured-alloy-anode-big-revolutionizing.html

Battery related breakthroughs are necessary as global energy needs are increasing, but the development of next-generation electrochemical energy storage systems with high energy density and acceptable cycling life are problematic.

 

Biotech

Rare quadruple-helix DNA found in living human cells with glowing probes

https://phys.org/news/2021-01-rare-quadruple-helix-dna-human-cells.html

G-quadruplexes have been in higher concentrations in cancer cells, so they may be linked to the disease, which shows us how ignorant we still are about cancer and dna biochemistry.

 

Scientists develop a cheaper method that might help create fuels from plants

https://phys.org/news/2021-01-scientists-cheaper-method-fuels.html

Creating biofuels and other bioproducts more cheaply is becoming even more economically competitive and sustainable due to this discovery which can help investments in the existing conventional car fleet become more eco-friendly and extend their lifetime.

 

Covid: Swansea Uni develops ‘world’s first’ vaccine smart patch – BBC News

https://www.bbc.com/news/uk-wales-55548670

An invention that will be demonized by conspiracy theorists, but still very useful and in the far future, people will wonder how we could live without it.

 

France’s Carmat Soars as Artificial Heart Hopes Become Reality – Bloomberg

https://www.bloomberg.com/news/articles/2020-12-24/france-s-carmat-soars-as-artificial-heart-hopes-become-reality

Given that heart disease is the leading death cause, artificial hearts could be ubiquitous among elderly people in a few decades.

 

Machine learning

Machine learning at the speed of light: New paper demonstrates use of photonic structures for AI

https://techxplore.com/news/2021-01-machine-paper-photonic-ai.html

While traditional electronics, which usually work in the low GHz range, this optical modulation could operate within 50 to 100 GHz range and make AI training a thousand or even million times more efficient that the state of the art systems.

 

AutoML-Zero. It took us 60 years to discover a thing… | by Sahil Uppal | Towards Data Science

https://towardsdatascience.com/automl-zero-b2e065170941?gi=660d72152ede

The possibilities of AutoML-Zero are endless, as it demonstrates how machine learning can come up with mathematical methods on its own and gives us a preview of what’s coming within the next 50 years.

 

DALL·E: Creating Images from Text

https://openai.com/blog/dall-e/?s=03

DALL·E is an impressive 12-billion parameter version of GPT-3 trained to produce images from text descriptions, based on a dataset of text–image pairs coming up with curious anthropomorphized versions of animals and objects, mixing unrelated concepts in plausible ways, generating text, and applying transformations to existing images showing that AI can exhibit a vivid imagination.

 

New Quantum Algorithms Finally Crack Nonlinear Equations

https://www.quantamagazine.org/new-quantum-algorithms-finally-crack-nonlinear-equations-20210105/

Novel, concrete examples of quantum algorithms are part of the second era of computing emerging from quantum computing.

 

Materials

Wood-inspired cement with high strength and multifunctionality

https://phys.org/news/2021-01-wood-inspired-cement-high-strength-multifunctionality.html

Biomimetic man-made materials and AI inventions are a recurring pattern during the last decades; even cement can thus be enhanced and further optimized.

 

Computing

80 Times More Energy Prototype Superconducting Chips that Will Soon Reach 10 Ghz | NextBigFuture.com

https://www.nextbigfuture.com/2020/12/80-times-more-energy-prototype-superconducting-chips-that-will-soon-reach-10-ghz.html

At 4.2 Kelvin, these chips will only operate under extreme laboratory conditions, however, material science discoveries could make them more mainstream, very soon.