Inside Updates

Near the end of the last century, Bill Gates saw the prospect of unifying citizens of nearly 200 countries, speaking more than 7,000 languages, coming together in common dialogue through the suddenly burgeoning web community.

“The Internet is becoming the town square for the global village of tomorrow,” he declared.

The Internet certainly has since drawn the world closer and has enriched global communications, commerce, research and entertainment immeasurably.

But a recent report reminds us—as if we really needed reminding—that along with progress sometimes come problems.

Researchers at Amazon Web Services Artificial Intelligence Lab and the University of California, Santa Barbara, say that after examining more than 6 billion sentences across the web, they have found that more than half had been translated into two or more different languages. The translations, they found, were often poor. And with each successive translation into other languages, some up to eight or nine, the results became worse.

The report, “A Shocking Amount of the Web is Machine Translated: Insights from Multi-Way Parallelism,” was uploaded to the preprint server arXiv Jan. 11.

“The low quality of these … translations indicates they were likely created using machine translation,” the authors report. “Our work raises serious concerns about training models such as multilingual large language models on both monolingual and bilingual data scraped from the web.”

The researchers said texts are not only being translated by artificial intelligence but are being created by AI as well. They observed rates of AI-generated translations were highest among lower-resource languages, such as Wolof and Xhosa, African languages.

“We find that highly multi-way parallel translations are significantly lower quality than two-way parallel translations,” the authors continue.

That means that as trillions of bits of data are ingested for AI training operations, regions under-represented on the web, such as African nations and other countries with more obscure languages, will face greater challenges in establishing reliable—and grammatical—large language models. With few native resources to draw upon, they must heavily rely on tainted translations flooding the market.

Mehak Dhaliwal, a former applied science intern at Amazon Web Services, told Motherboard in an interview, “We actually got interested in this topic because several colleagues who work in machine training and are native speakers of low resource languages noted that much of the internet in their native language appeared to be machine training generated… Everyone should be cognizant that content they view on the web may have been generated by a machine.”

The Amazon researchers found bias in selection of content used for AI training.

They state, “Machine generated, multi-way parallel translations not only dominate the total amount of translated content on the web in lower resource languages, it also constitutes a large fraction of the total web content in those languages.”

Such content, they suggested, tends to be simpler, lower-quality passages “likely produced to generate ad revenue.” Since fluency and accuracy are lower for machine-trained material, numerous translations will lead to even less accurate content and increase the odds of AI hallucination.

Sometimes, computer-generated translations over the years have led to unintentionally humorous or embarrassing interpretations.

Google misinterpreted a phrase “Russia is a great country” and referred instead to Mordor, a fictional village in J.R.R. Tolkien’s “The Lord of the Rings.” Facebook’s translation software in 2019 inadvertently referred to China’s President Xi Jinping as “Mr. S***hole” several times in an English article translated from Burmese text. Facebook immediately apologized and blamed the mishap on a “technical error.”

And a medical prescription translation tool for Armenian speakers provided some unfortunate advice for a patient with a headache.

English: “You can take over-the-counter ibuprofen as needed for pain.”

Translation to Armenian: “You may take anti-tank missile as much as you need for pain.”

More information:
Brian Thompson et al, A Shocking Amount of the Web is Machine Translated: Insights from Multi-Way Parallelism, arXiv (2024). DOI: 10.48550/arxiv.2401.05749

Journal information:
arXiv

Citation:
Faulty machine translations litter the web (2024, January 22)
retrieved 27 June 2024
from https://techxplore.com/news/2024-01-faulty-machine-litter-web.html

This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no
part may be reproduced without the written permission. The content is provided for information purposes only.

Source link

Earth scientist studies why U.S. has so many tornadoes

June 27, 2024

Across the Midwest during the warmer months, studying the sky for signs of storms and tornadoes becomes one of the most popular pastimes.

Dan Chavas, an associate professor in the Department of Earth, Atmospheric, and Planetary Sciences at Purdue University’s College of Science, takes it further: All day every day, he studies what makes tornadoes tick. Working at the intersection of climate science and meteorology, he looks at the big picture of what causes severe storms and tornadoes—and what dictates where they occur.

“I study both the climate and extreme weather,” Chavas says. “My research asks, “Why do we have severe thunderstorms or tornadoes at all?” There are specific regions on Earth that have more storms, more tornadoes than other places. What creates these stormy regions?”

The central and eastern regions of the United States are among the top spots for severe thunderstorms and form the hot spot for the Earth’s most damaging and frequent tornadoes. Chavas uses real-world computer models to conduct experiments to determine what contributes to the formation of these storms.

“We have had these decades-old assumptions about what causes storms,” he says. “We’re validating those hypotheses and figuring out what makes North America such a hot spot.”

Moving heaven and earth

Chavas isn’t a storm chaser. He’s not out there in a weather van topped with satellite wires hunting down individual storms for the insights they might yield. Nor can he grow storms in his lab or unleash tornadoes to understand their anatomy or behavior.

Instead, he harnesses decades of rich, detailed historical data and complex computer models to imagine and test what-if scenarios. He’s a storm tester.

“We use weather and climate models, as well as extensive databases of thunderstorms, lightning strikes, atmospheric data and more, to ask, “What if the world was different?'” Chavas says. “We can use these models as laboratories to ask questions like ‘What happens to the weather if you flatten the Rocky Mountains? What about if you fill in the Gulf of Mexico? What aspects of the modern continental and mountain configurations really matter? Let’s actually test this prevailing, conventional wisdom.'”

Both of those hypotheticals—flattening the Rockies and filling in the Gulf of Mexico—are the focus of studies Chavas and his team have conducted.

For more than 50 years, established wisdom said that the Gulf of Mexico, a source of warm, wet air flowing inland to the east of the Rocky Mountains, plays a major role in the formation of North America’s tornadoes. But no one knew for sure.

“It was a very reasonable hypothesis,” Chavas says. “There were a lot of very reasonable explanations. But no one had been able to test these 50-year-old ideas because they came about when there weren’t climate models with the necessary computational power. Now we can really start to understand the physics of the situation.”

When his team virtually filled in the Gulf of Mexico with land, they found that a dry Gulf of Mexico affected the frequency and severity of storms far less than they had expected. Without the Gulf of Mexico, severe thunderstorms shifted eastward from the central Great Plains into Illinois, although they were reduced over southern Texas.

“Severe thunderstorms and tornadoes form in environments with specific ingredients for how temperature, moisture, and especially wind speed and direction change with height in the atmosphere,” Chavas says.

“The climate determines where and when those ingredients can be found together to produce these types of storms. Computer models let us understand why the ingredients are there in the first place and what role they each play in the weather we see.”

In his most recent study with graduate student Funing Li, published in the Proceedings of the National Academy of Sciences, the team compared severe weather potential in North America, famous for tornadoes, with South America, which has a geography similar to North America’s and also many severe thunderstorms, but far fewer tornadoes.

They found that the rough texture of the land surface east of the Andes mountains, its roughness determined in part by the hills and tall trees of the Amazon region, may play a large role in preventing tornadoes over central South America. In contrast, in North America many tornadoes form east of the Rockies, where air flows in from the much smoother ocean surface of the Gulf of Mexico.

The team first used climate model experiments in which equatorial South America was smoothed to be similar to an ocean surface, which drastically increased central South America’s tornado potential. They also performed experiments in which the Gulf of Mexico region was roughened to be similar to a forested land surface, which strongly suppressed North American tornado potential.

“A rough surface upstream means that downstream the wind is no longer changing speed and direction with height very strongly near the surface, which we refer to as ‘wind shear,'” Chavas says. “It doesn’t change ingredients for severe thunderstorms, but the wind shear in the 1 kilometer of air above the ground is a critical ingredient for tornadoes.”

Storm warning

Real weather and real-world applications fascinate Chavas, a fascination born after a storm-torn tree fell on his house in Wisconsin when he was 4 years old.

The real-world implications of his research—what will the weather be like next week, next month, next year and next century—are what drives him.

“If we want to understand how climate change will affect weather in the future, we need to understand how climate determines weather in the first place,” Chavas says. “We don’t have a very good understanding of how climate controls the severe weather we have.”

Understanding how surface roughness and land use changes weather, for example, may enable future humans to better predict—and even partially affect—weather patterns. If the rough land of the Amazon, including a component from the trees of the Amazon, protects South America from tornadoes, could the regrowth of the United States’ eastern forests affect tornadoes, too?

Climate change affects the flow patterns of the atmosphere and moisture distribution on land, Chavas says.

“If we change the land surface and the trajectory of air flowing inland from the Gulf of Mexico, it may have a direct impact on these ingredients that give rise to tornadoes farther inland. When we think about climate change, we think about it getting hotter and the land getting drier.

“But if the jet stream changes where and how quickly air flows inland, it can change where and how tornadoes form. Places that didn’t see them before may see them more, and places that had more may see fewer,” he says. “We need to understand the weather now to help us better predict the weather of the future.”

More information:
Funing Li et al, Upstream surface roughness and terrain are strong drivers of contrast in tornado potential between North and South America, Proceedings of the National Academy of Sciences (2024). DOI: 10.1073/pnas.2315425121

Provided by
Purdue University

Citation:
New twists on tornadoes: Earth scientist studies why U.S. has so many tornadoes (2024, June 27)
retrieved 27 June 2024
from https://phys.org/news/2024-06-tornadoes-earth-scientist.html

Source link

Study projects loss of brown macroalgae and seagrasses with global environmental change

June 27, 2024

Projected loss of brown macroalgae and seagrasses with global environmental change — Present distribution and projected end-of-century changes in global macrophyte species diversity. Credit: *Nature Communications* (2024). DOI: 10.1038/s41467-024-48273-6

Researchers predict that climate change will drive a substantial redistribution of brown seaweeds and seagrasses at the global scale. The projected changes are alarming due to the fundamental role of seaweeds and seagrasses in coastal ecosystems, and provide evidence of the pervasive impacts of climate change on marine life.

In a collaborative study between the University of Helsinki and the EU Joint Research Centre, researchers for the first time have modeled the future distribution of brown seaweeds and seagrasses at the global scale. They predict that by 2100, climate change will drive a substantial redistribution of both groups globally: Their local diversity will decline by 3–4% on average and their current distribution will shrink by 5–6%. More notably, the preferred habitat for both brown seaweeds and seagrasses will undergo a substantial global reduction (78–96%) and will shift among marine regions, with potential expansions into Arctic and Antarctic regions.

The research is published in the journal Nature Communications.

“We find it alarming that coastal areas worldwide will become dramatically less hospitable for habitat-forming macrophytes, as this might have severe and widespread impacts on coastal ecosystem functioning at the global scale. Interestingly, while global percentual declines in diversity show similar trends for seagrasses and brown macroalgae, the regional patterns are strikingly different between the two groups,” says Federica Manca, the lead author of the study from the University of Helsinki.

Why should we care about seaweeds and seagrasses?

Brown seaweeds and seagrasses provide important ecological and socio-economic services in coastal areas worldwide: They support coastal biodiversity and fisheries, ensure coastal protection, participate in ocean nutrient recycling, contribute to carbon sequestration and climate change mitigation.

As climate change is severely threatening macrophyte habitats and the services they provide, we urgently need to understand how both brown seaweeds and seagrasses will respond to changing climatic conditions in the coming decades.

Previous studies have modeled the future distribution of these habitat-forming macrophytes, focusing on regional or local scales only and on a limited number of species. In contrast, this study is the first to provide a comprehensive view of the effects of climate change on more than 200 species of brown seaweeds and seagrasses at the global scale.

The results show that the redistribution of these habitat-forming marine macrophytes will be geographically heterogeneous, and highlight the regions where the loss of macrophyte diversity and habitat will be most severe, such as the Pacific coast of South America for brown seaweeds, and the coast of Australia for seagrasses. Additionally, researchers have identified macrophyte species that will be more severely affected by climate change, like the Atlantic seaweed Laminaria digitata. The findings can help identify target areas and species for conservation, potentially buffering the impact of climate change.

Surprisingly, and contrary to expectations, the models did not predict severe losses of brown seaweed or seagrass diversity in the tropics but rather at intermediate and high latitudes, such as along the Atlantic coasts of Europe and in the Baltic Sea. This indicates that end-of-century climatic conditions in these regions might exceed the tolerance limits of resident macrophyte species. The Baltic Sea is at the forefront in the rate at which climate change is influencing the ecosystem.

“Combined with a legacy of multiple other disturbances (such as eutrophication) and low species diversity with only a few brown seaweeds and seagrasses, the Baltic Sea is exceptionally vulnerable to these predicted changes,” says Alf Norkko, professor at the Tvärminne Zoological Station, University of Helsinki.

“Another surprising—and alarming—result is the dramatic loss of highly suitable habitat for both macroalgae and seagrasses globally: Coastal areas worldwide will become substantially less hospitable for habitat-forming macrophytes,” adds Dr. Mar Cabeza from the Global Change and Conservation Group at the University of Helsinki.

The disappearance of these habitat-forming macrophytes can trigger cascading effects on other species, compromising the integrity of entire ecosystems and undermining ecological and socio-economic services important to human society. Thus, forecasting changes in the distribution of habitat-forming species is crucial to raise awareness of climate change impacts and foster conservation efforts accordingly.

“Our findings confirm, once again, that climate change might have profound impacts on ecosystems, promoting rapid and most often detrimental changes to the diversity and resilience of natural communities. In fact, habitat-forming macrophytes support biodiversity through an exceptional diversity of ecological interactions.

“Hence, their projected loss and redistribution might lead to unpredictable cascading effects, most likely resulting in the local extinction of many associated species,” says Giovanni Strona from the EU Joint Research Centre.

More information:
Federica Manca et al, Projected loss of brown macroalgae and seagrasses with global environmental change, Nature Communications (2024). DOI: 10.1038/s41467-024-48273-6

Provided by
University of Helsinki

Citation:
Study projects loss of brown macroalgae and seagrasses with global environmental change (2024, June 27)
retrieved 27 June 2024
from https://phys.org/news/2024-06-loss-brown-macroalgae-seagrasses-global.html

Source link

‘Splitting Germany into several price zones on the day-ahead power market might not achieve the intended goals’

June 27, 2024

electricity — Credit: Unsplash/CC0 Public Domain

Temporal and locational variations in power generation result in relatively large fluctuations in the power supply and power prices. Grid infrastructure providers are increasingly forced to intervene in order to balance supply and demand and avoid power outages. Consequently, the EU Commission is assessing a possible split of the German uniform price zone and the day-ahead market into smaller price zones.

Researchers at the Technical University of Munich (TUM) have shown that smaller price zones would have little effect on the power price or re-dispatch measures. By contrast, the use of nodal pricing would reduce overall costs for providing energy by 9%. The study is published in the journal Operations Research.

What political discussions are taking place on this topic, and what does your research show?

Current discussions revolve around dividing Germany into two to four bidding zones instead of the current uniform zone. The EU Commission has ordered a bidding zone review to reassess the price zones in the EU. In our study, we used the data set from the bidding zone review to analyze trends in power prices and the costs of re-dispatch measures under the assumption that Germany was divided into the proposed bidding zones. Never before was such a comprehensive data set available for analysis.

In addition, we calculated local (nodal) prices. Under a zonal price system, a single hourly price applies for the entire bidding zone. By contrast, a nodal system sets an individual price for each node.

How much would a nodal pricing rule reduce the total power costs?

With power price zones, it was evident that there would be practically no difference between the individual zonal prices. At the same time, the price variance and re-dispatch costs would not decrease significantly compared to the power price in a German uniform price zone. Several experts had expected a stronger impact from a zonal split. However, we do not see this in the data set provided for the bidding zone review.

With our calculations we were able to show that the lowest total costs would result if Germany were to use nodal pricing. Compared to the uniform price or zonal pricing, the total costs would be around 9% lower. That is in particular due to the fact that the market mechanism takes network restrictions into account and that it succeeds in efficiently allocating the available resources. As a result, costly re-dispatch measures are largely avoided.

To understand these effects, one must consider how power pricing currently works in Germany.

Power prices across Europe are determined in a day-ahead auction. Across Europe there are various price zones, each with its own electricity prices. The zones can be entire countries. However, some countries, such as Italy, are divided into several zones.

What is the situation in Germany?

Germany has only one price zone. It may happen, however, that large amounts of wind energy are generated in northern Germany, for example, while demand is particularly high in the south. But due to the limited grid capacity, not enough power can be transmitted from the north to the south. The uniform price for Germany determined in the day-ahead auction does not take this into account.

What problems result from the uniform price?

Under the current pricing mechanism, there is little incentive to adjust consumption on the demand side when electricity is in short supply—because the price at locations where power is scarce is the same everywhere in Germany.

In the example I mentioned, the production of wind power needs to be throttled in the north while expensive gas-fired power stations are fired up in the south to cover demand there. These re-dispatch measures are very costly. In 2023 they totaled 3.1 billion euros. That amount was passed on to consumers.

How could that be changed?

By replacing uniform prices for large power pricing zones by locational prices set for individual nodes in the grid. This system operates in many countries around the world, including the U.S. Texas has more than 4000 nodes, for example. Due to the fluctuating power supply, some nodes have excess supply at some times and less at others. This results in temporary downward or upward price movements at those locations.

When prices rise, there would be an incentive for the industry to reduce demand at these nodes. Either production would be shifted elsewhere or energy storage would be utilized. The demand-side flexibility would help grid operators to stabilize the network and substantially reduce the need for re-dispatch measures.

More information:
Mete Şeref Ahunbay et al, Pricing Optimal Outcomes in Coupled and Non-convex Markets: Theory and Applications to Electricity Markets, Operations Research (2024). DOI: 10.1287/opre.2023.0401

Provided by
Technical University Munich

Citation:
‘Splitting Germany into several price zones on the day-ahead power market might not achieve the intended goals’ (2024, June 27)
retrieved 27 June 2024
from https://phys.org/news/2024-06-germany-price-zones-day-power.html

Source link

Researchers successfully develop domestic 6G antenna measurement system

June 27, 2024

KRISS successfully develops domestic 6G antenna measurement system — Mobile 6G antenna measurement system developed by KRISS. Credit: Korea Research Institute of Standards and Science (KRISS)

In April 2019, South Korea ambitiously launched the world’s first 5G mobile communication service. While 5G in the 3.5 GHz band was commercialized, the communication quality did not meet consumer expectations. The installation of base stations in the 28 GHz band, which would provide true 5G service, was slow due to profitability concerns.

Consequently, the government reclaimed the frequency bands from all three major telecommunications companies last year. As countries around the world prepare for the 6G era, it is time to reflect on the disappointing experiences of 5G commercialization and focus on building the 6G infrastructure.

Korea Research Institute of Standards and Science (KRISS) has succeeded in developing domestically produced equipment to evaluate the performance of 6G communication antennas.

The paper is published in the journal IEEE Transactions on Instrumentation and Measurement.

As the frequency band increases, communication speed typically improves, but the communication range shortens. Since 6G communication (planned for 7–24 GHz) has a higher frequency band than the current 5G communication (3.5 GHz), antenna-related technologies are needed to address the issue of shortened communication range.

To ensure the proper functioning of these advanced 6G antennas, accurate performance evaluation is essential. Precise performance measurements can help identify and correct causes of malfunction in the prototype stage, improve quality, and shorten the time to mass production.

The research team from the KRISS Electromagnetic Wave Metrology Group has developed a 6G antenna measurement system based on a non-metallic sensor using an optical method.

To evaluate antenna performance, the sensor is placed at a certain distance to measure the electromagnetic waves generated by the antenna. Previously, metallic sensors were used. This caused coupling effects due to the electromagnetic wave reflection properties of metal, resulting in distorted measurements. This problem was easily resolved by replacing them with non-metallic sensors the size of a grain of rice.

The distance between the sensor and the antenna during measurements has decreased from several meters to a few millimeters, with measurement time reduced by more than one-tenth. Moreover, unlike previous measurements that required very large, fixed facilities such as anechoic chambers, the measurement equipment developed by KRISS is lightweight, similar in size and weight to a computer tower, making it portable and suitable for use in standard laboratories.

KRISS has transferred this technology to East Photonics Co., Ltd., a company specializing in fiber optic communication and repeaters, for a royalty of KRW 300 million, and a signing ceremony was held on April 8 at the KRISS administrative building.

Young-Pyo Hong, a principal researcher at KRISS, stated, “Currently, domestic research related to 6G is concentrated only in the materials and components fields, and studies have yet to be conducted on measurement equipment. Learning from the disappointing experiences with 28 GHz 5G communication, we plan to prioritize the establishment of 6G infrastructure, with the development of measurement equipment being a crucial part.”

Ho-Joon Seok, president and CEO of East Photonics Co., Ltd., said, “All smartphone and base station antenna measurement equipment is expensive and foreign-made, but commencing now we will take the lead in domesticating 6G antenna measurement equipment in close collaboration with KRISS. Unlike existing measurement equipment, our lightweight and mobile measurement equipment will be a strong point as we steadily plan for commercialization.”

More information:
Dong-Joon Lee et al, Integrated Electrooptic Sensor for Intense Electromagnetic Pulse Measurements, IEEE Transactions on Instrumentation and Measurement (2023). DOI: 10.1109/TIM.2023.3284921

Provided by
National Research Council of Science and Technology

Citation:
Researchers successfully develop domestic 6G antenna measurement system (2024, June 27)
retrieved 27 June 2024
from https://techxplore.com/news/2024-06-successfully-domestic-6g-antenna.html

Source link

1...1,0711,0721,073...1,162 Page 1,072 of 1,162

24/7 News

Science
World News
Health
Entertainment and arts
Top Stories
Technology
Business
Sports

About us

Ideal Inside Aims at showcase different blog/News storiesfrom different regions all around the world along with other benefits to have you experienc at best edge.

Facebook

Instagram

Youtube

I've read and accept the Privacy Policy.

Inside Updates

Subscribe to newsletter

Movies

TV Shows

Music

Celebrity

Scandals

Drama

Lifestyle

Health

Technology

Company

Moving heaven and earth

Storm warning

Why should we care about seaweeds and seagrasses?

What political discussions are taking place on this topic, and what does your research show?

How much would a nodal pricing rule reduce the total power costs?

What is the situation in Germany?

What problems result from the uniform price?

How could that be changed?

24/7 News

About us

Latest Articles

Most Popular

Subscribe