Saturday Quotes: Bacterial Warfare, Self-Programming Language Model, Passive Cooling in the Big City

There’s a lot of science news in seven days, so just because a new study isn’t cited here Saturday morning doesn’t mean it didn’t happen. A plot many other things happened. But also discover these four stories:

Bacteria use phages as weapons

In agricultural areas located on cultivated land, a variant of the bacteria Pseudomonas viridiflava spreads and becomes the dominant microbe, but this does not occur on uncultivated land. Researchers at the University of Utah wanted to understand why, but early in their study they observed something so strange and unexpected that it reoriented the entire study.

By studying the genome of bacterial pathogens, they discovered that one specimen had captured a phage – a virus that attacks bacteria – and reused it to kill its own bacterial competitors. Specifically, bacteria have acquired non-self-replicating pools of repurposed phages called tailocins, which penetrate the outer membranes of pathogens and kill them.

Speculating on this finding, lead author Talia Backman says that tailocins could potentially lead to new antibiotics to solve the antimicrobial resistance crisis: “Although tailocins have already been found in other bacterial genomes and have been studied in the laboratory, their impact and their evolution in nature The bacterial populations were not known. The fact that we found that these wild plant pathogens all have tailocins and that these tailocins evolve to kill neighboring bacteria shows how important they can be in nature.

The language model programs itself to make sense

Large language models are capable of producing text. I was going to complete this thought with “it’s compelling and convincing” or something like that, but I had to stop and lower my head to the table because the AI ​​hype cycle broke its bond and became completely detached from reality and I am fatigue.

Existing LLMs are nothing more than text predictors without any knowledge of semantics or logic, and therefore can only help humans with the generation of text containing zero percent symbolic reasoning, and they will now be present in all devices. Okay, I’m not here to complain about the AI ​​hype.

In an effort to improve the performance of LLMs, MIT researchers have proposed an innovative technique to perform natural language, mathematics, and data analysis tasks by generating Python code. They call this approach natural language embedded programming and report 90% accuracy on a wide range of reasoning tasks.

The technique is a four-step process. In the first step, the NLEP calls the packages required to perform a task. Secondly, it imports natural language representations of the knowledge or data required by the task. In the third step, the model generates a function that calculates the answer. And in the fourth step, the model generates the result in natural language. It is also more effective for certain tasks, notably those in which a user has many similar questions; Instead of generating a new Python program for each query, NLEP can generate a main program and modify variables for each query.

Cool shirt

As summer heat domes take hold in regions across the Northern Hemisphere, researchers at the UChicago Pritzker School of Molecular Engineering report a new wearable fabric that can protect city dwellers in the specific conditions of urban heat islands . Existing cooling fabrics work by dispersing direct sunlight. But sunlight is visible while thermal radiation from building materials, sidewalks and infrastructure is infrared.

The engineers sought to create a textile with two optical properties that could reflect both sources. And because it operates passively, it can have cooling applications in areas where energy consumption is increasing. Beyond heat-reflecting clothing, the material has potential as a building material to reduce indoor temperatures or as insulation for cars. The researchers also suggest it could be used to transport perishable foods, reducing demand for active refrigeration systems.

Immature children, according to a study

Researchers from the National University of Singapore report that a lower ratio of neuronal excitation (E) to neuronal inhibition (I) is a positive sign of brain maturation. They found that children with a lower E/I ratio performed better in school and on cognitive tests. . Previous studies have shown that excessive excitation or inhibition leads to a higher risk of brain disorders, including autism, Alzheimer’s disease and others. The study explored the development of E/I in youth by analyzing brain MRIs and cognitive test results of 885 children, adolescents and young adults.

In particular, the team created a non-invasive technique to study neuronal excitation and inhibition responses. In the first part of the experiment, subjects ingested the anti-anxiety drug Xanax or a placebo before a brain MRI. This allowed researchers to establish that Xanax increases neuronal inhibition, such that the overall E/I ratio decreases. In the second part of the experiment, the team established the link between E/I ratio and cognitive function by administering cognitive tests. Participants with lower E/I outperformed those with higher ratios.