Digital Tool Helps Understand The Past, Predict Evolution Of The Earth’s Surface

Most detailed geological model reveals Newswise — Climate, tectonics and time combine to create powerful forces that craft the face of our planet. Add the gradual sculpting of the Earth’s surface by rivers and what to us seems solid as rock is constantly changing.

However, our understanding of this dynamic process has at best been patchy.

Scientists today have published new research revealing a detailed and dynamic model of the Earth’s surface over the past 100 million years.

Working with scientists in France, University of Sydney geoscientists have published this new model in the prestigious journal Science.

For the first time, it provides a high-resolution understanding of how today’s geophysical landscapes were created and how millions of tonnes of sediment have flowed to the oceans.

Picture : TheUNN

Lead author Dr Tristan Salles from the University of Sydney School of Geosciences, said: “To predict the future, we must understand the past. But our geological models have only provided a fragmented understanding of how our planet’s recent physical features formed.

“If you look for a continuous model of the interplay between river basins, global-scale erosion and sediment deposition at high resolution for the past 100 million years, it just doesn’t exist.

“So, this is a big advance. It’s not only a tool to help us investigate the past but will help scientists understand and predict the future, as well.”

Using a framework incorporating geodynamics, tectonic and climatic forces with surface processes, the scientific team has presented a new dynamic model of the past 100 million years at high resolution (down to 10 kilometres), broken into frames of a million years.

Second author Dr Laurent Husson from Institut des Sciences de la Terre in Grenoble, France, said: “This unprecedented high-resolution model of Earth’s recent past will equip geoscientists with a more complete and dynamic understanding of the Earth’s surface.

“Critically, it captures the dynamics of sediment transfer from the land to oceans in a way we have not previously been able to.”

Dr Salles said that understanding the flow of terrestrial sediment to marine environments is vital to comprehend present-day ocean chemistry.

“Given that ocean chemistry is changing rapidly due to human-induced climate change, having a more complete picture can assist our understanding of marine environments,” he said.

The model will allow scientists to test different theories as to how the Earth’s surface will respond to changing climate and tectonic forces.

Further, the research provides an improved model to understand how the transportation of Earth sediment regulates the planet’s carbon cycle over millions of years.

“Our findings will provide a dynamic and detailed background for scientists in other fields to prepare and test hypotheses, such as in biochemical cycles or in biological evolution.”

Authors Dr Salles, Dr Claire Mallard and PhD student Beatriz Hadler Boggiani are members of the EarthColab Group and Associate Professor Patrice Rey and Dr Sabin Zahirovic are part of the EarthByte Group. Both groups are in the School of Geosciences at the University of Sydney.

The research was undertaken in collaboration with French geoscientists from CNRS, France, Université Lyon and ENS Paris.

Your Next Colleague Might Be Artificial Intelligence

Newswise — Most companies are likely to ‘employ’ a digital human within a decade, according to researchers from the University of Sydney Business School.

A digital human is a combination of artificial intelligence and a computer-generated avatar that can interact with people and produce realistic human speech and facial expressions in real time.

From Siri to ChatGPT, most of us have already interacted with artificial intelligence – sometimes without even knowing. Digital humans take this technology to the next level by overlaying it with the kind of powerful and lifelike visual effects once limited to Hollywood blockbusters.

Writing in Harvard Business Review, researchers from the University of Sydney, Indiana University and Iowa State University collectively drew on their research to offer insights on the future of the workforce where humans work side-by-side with digital employees.

They write: “When deployed at scale, digital humans will radically change the business landscape. They may not be as capable or versatile as human employees, but they have clear advantages when it comes to cost, customisability, and scalability.

“Once ‘hired’, they never tire, never complain, never seek a raise, and always follow company policy.”

What can a digital employee do?

Article co-author Dr Mike Seymour, senior lecturer at the University of Sydney and co-director of Motus Lab, said digital humans have already been successfully employed in a variety of fields.

The search for more engaging customer experiences has seen the invention of digital sales representatives who can suggest and model clothing and cosmetics for online shoppers.

Digital humans can be grouped into four basic categories:

Virtual agent: serves multiple users to complete specific, one-time tasks, such as customer service.

Virtual assistant: supports the user regularly with a specific task such as shopping or physical therapy.

Virtual influencer: supply their followers with non-customised content to market products or experiences.

Virtual companion: develops a close relationship with the user that is based on interaction for its own sake, not the completion of specific tasks, but can assist with tasks such as reminding users to take their medication.

“While I would cautiously warn companies not to blindly dive into this new technology, there are productivity gains in pairing the correct type of digital employee with suitable business functions,” Dr Seymour said.

“A traditional text interface such as a chatbot is preferred for quick interactions, but a digital human can do a much better job communicating complex instructions and emotionally engaging with a customer. They also work well in a situation where the customer isn’t sure what they want and is open to exploring options,” adds co-author Dan Lovallo, Professor of Strategy, Innovation and Decision Sciences at the University of Sydney.

Human employees will still be vital

Despite the promise and broad applications of AI-driven technology, the researchers do not believe digital humans will replace the real thing in any industry.

“We’re still a long way, if ever, from AI being able to recognise and respond to all the complex nuances of verbal and non-verbal communication,” said co-author Professor Kai Riemer, Head of Business Information Systems at the University of Sydney Business School.

“As realistic and lifelike as a digital human may seem, ultimately it is never ‘thinking’ and lacks any true understanding of what is being communicated.

“Its greatest advantage over existing AI technology is also its ultimate limitation: our own innate desire for face-to-face human interaction.”

Join the authors of the Harvard Business Review article in a webinar at 8:30am AEDT on Friday, February 25, where they will share their research on digital humans. The discussion will include more use cases of digital employees across multiple industries, current state-of-the-art strategies for deploying digital employees, and their thoughts on how digital employees will shape the landscape of future workforce.

How Does Our Immune System Reacts to COVID-19 Variants

Newswise — Australian scientists researching how our immune system responds to COVID-19 have revealed that those infected by early variants in 2020 produced sustained antibodies, however, these antibodies are not as effective against contemporary variants of the virus.

The research is one of the world’s most comprehensive studies of the immune response against COVID-19 infection. It suggests vaccination is more effective than the body’s natural immune response following infection and shows the need to invest in new vaccine designs to keep pace with emerging COVID variants.  Published today in PLOS Medicine, the study was made possible by a partnership between the University of Sydney, Kids Research, Sydney Children’s Hospitals Network, the Kirby Institute at UNSW Sydney, Australian Red Cross Lifeblood, St Vincent’s Hospital and NSW Health Pathology, as well as other local and international collaborators.

The team analysed the serum of 233 individuals diagnosed with COVID-19 over 7 months and uncovered that the level of immunity over time is dependent on disease severity and the viral variant. They show that antibodies developed during the first wave had reduced effectiveness against six variants, ranging from those observed in the second wave in Australia through to three variants of concern that have driven the global pandemic in the UK, Brazil and South Africa.

How do we study the immune response?

The serum of COVID-19 infected individuals was of interest as it is the part of our blood that contains crucial information about our immune system. Analysis of the serum made it possible to create a detailed timeline of the level of ‘neutralising antibodies’ produced against COVID-19 infection, and so to see if there was long-term immunity. 

Neutralising antibodies are part of our immune system’s frontline arsenal that is triggered during infection and vaccination. Their job is to shield cells that are usually the target of a pathogen (such as the SARS-CoV-2 virus which causes the COVID-19 disease) from being infected. The level of neutralising antibody response can be a defining feature of how effectively our body fights off illness.

Interestingly, a rare group of ‘super responders’ was also identified as an exception.  This group of ‘super responders’ had a stable and robust level of antibodies across all COVID-19 variants. The researchers say this group could prove useful for investigating the potential of convalescent plasma (using blood from people who have recovered to treat others) which has so far proven ineffective against severe COVID-19 illness. In addition, key donors could be looked at closely and their antibodies cloned for future therapeutic use.

Why it is important?

Co-senior author Associate Professor FabienneBrilot of the University of Sydney and Kids Research, Sydney Children’s Hospitals Network, and her research team led the analysis branch of the study, using highly sensitive tools they developed to study the antibodies in detail.

“We can learn a great deal from these people who were infected in the first wave in Australia as they were infected with the same variant that our current vaccines are based on,” said Associate Professor Brilot.  “While the approved vaccines are showing good responses, our study highlights the importance of continued vaccine development, especially taking into account the differences in variants.” Co-senior author Associate Professor Stuart Turville of the Kirby Institute said the study was conducted to investigate the level, breadth and longevity of the immunity generated from COVID-19 infection and whether mutation of the virus compromises immunity.

“What this work has shown us is that current observations about vaccines show they offer a much broader protection against COVID-19 and its variants than the body’s natural immune response following infection, which is usually only protective against the variant of the virus that the person was infected with. We, therefore, should not rely on the body’s natural immune response to control this pandemic, but rather the broadly protective vaccines that are available.”

Key findings 

  • SARS-CoV-2 antibody responses are sustained for up to seven months post-infection.
  • The immune response remained stable in some individuals, and while it decreased in others, no individual showed a negative response during the seven-month period.
  • Levels of virus-neutralising antibodies were associated with COVID-19 severity.
  • Antibodies generated after early infection displayed a significantly reduced antibody binding and neutralisation potency to globally emerging viral variants.