At what stage should people be given the power to overrule AI in the workplace? New research suggests sooner is better.
Giving employees the power to overrule AI decisions from day one boosts motivation and enhances learning, according to a new study.
“Imagine you’re a financial specialist at a bank,” says Business School lecturer Dr. Frank Ma. “You input details for a mortgage application, and the AI system recommends declining it. While the system is based on hard data, as a human, you can recognize nuances—’soft’ information—that AI can miss. This is where the ability to overrule the system is beneficial.”
Dr. Ma, along with researchers Stijn Masschelein and Vincent Chong from the University of Western Australia, conducted a series of online tasks with 161 participants. The tasks were designed to simulate real-world scenarios in sectors where AI is utilized in decision-making.
Their findings suggest that when employees are empowered to overrule AI decisions early on, they become more motivated and better equipped to understand complex tasks.
The study also looked at the impact of incentive schemes, finding that combining early autonomy and incentive pay creates an environment where employees are more engaged and learn faster.
“Whether and when to override AI decisions is already a big issue in industries including banking and manufacturing, and it’s going to become one in many others that use algorithms for managerial decision-making,” says Ma.
“Overall, our study shows that giving employees the power to have the final say over AI early on is critical to their learning. Humans can pick up on nuances that artificial intelligence can’t, so people need the power to make the final call.”
Giving employees immediate flexibility, he says, provides more opportunities to override system decisions, and incentive pay ensures that employees put more effort into making the final call accurately.
“Employees with incentive schemes and immediate flexibility get a better understanding of their roles and improve their performance. We believe this is due to developing a more in-depth understanding of how the computer system or AI generates its decision.”
The study also shows that delaying this autonomy has a detrimental effect on employee motivation, potentially limiting learning and performance.
The working paper, “Incentive contracts and the timing to introduce flexibility on employee learning,” won the best paper award (management accounting) at the Accounting and Finance Association of Australia and New Zealand conference 2024.
Citation:
Early autonomy over AI boosts employee motivation, researchers suggest (2024, September 16)
retrieved 16 September 2024
from https://phys.org/news/2024-09-early-autonomy-ai-boosts-employee.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.
Most of us aren’t spending our days watching our houseplants grow. We see their signs of life only occasionally—a new leaf unfurled, a stem leaning toward the window.
But in the summer of 1863, Charles Darwin lay ill in bed, with nothing to do but watch his plants so closely that he could detect their small movements to and fro. The tendrils from his cucumber plants swept in circles until they encountered a stick, which they proceeded to twine around.
“I am getting very much amused by my tendrils,” he wrote.
This amusement blossomed into a decades-long fascination with the little-noticed world of plant movements. He compiled his detailed observations and experiments in an 1880 book called “The Power of Movement in Plants.”
A diagram tracking the circumnutation of a leaf over three days. Credit: Charles Darwin
In one study, he traced the motion of a carnation leaf every few hours over the course of three days, revealing an irregular looping, jagged path. The swoops of cucumber tendrils and the zags of carnation leaves are examples of inherent, ubiquitous plant movements called circumnutations—from the Latin circum, meaning circle, and nutare, meaning to nod.
Circumnutations vary in size, regularity and timescale across plant species. But their exact function remains unclear.
I’m a physicist interested in understanding collective behavior in living systems. Like Darwin, I’m captivated by circumnutations, since they may underlie more complex phenomena in groups of plants.
Sunflower patterns
A 2017 study revealed a fascinating observation that got my colleagues and me wondering about the role circumnutations could play in plant growth patterns. In this study, researchers found that sunflowers grown in a dense row naturally formed a near-perfect zigzag pattern, with each plant leaning away from the row in alternating directions.
This pattern allowed the plants to avoid shade from their neighbors and maximize their exposure to sunlight. These sunflowers flourished.
Researchers then planted some plants at the same density but constrained them so that they could grow only upright without leaning. These constrained plants produced less oil than the plants that could lean and get the maximum amount of sun.
While farmers can’t grow their sunflowers quite this close together due to the potential for disease spread, in the future they may be able to use these patterns to come up with new planting strategies.
Self-organization and randomness
This spontaneous pattern formation is a neat example of self-organization in nature. Self-organization refers to when initially disordered systems, such as a jungle of plants or a swarm of bees, achieve order without anything controlling them. Order emerges from the interactions between individual members of the system and their interactions with the environment.
Somewhat counterintuitively, noise—also called randomness—facilitates self-organization. Consider a colony of ants.
Ants secrete pheromones behind them as they crawl toward a food source. Other ants find this food source by following the pheromone trails, and they further reinforce the trail they took by secreting their own pheromones in turn. Over time, the ants converge on the best path to the food, and a single trail prevails.
But if a shorter path were to become possible, the ants would not necessarily find this path just by following the existing trail.
If a few ants were to randomly deviate from the trail, though, they might stumble onto the shorter path and create a new trail. So this randomness injects a spontaneous change into the ants’ system that allows them to explore alternative scenarios.
Eventually, more ants would follow the new trail, and soon the shorter path would prevail. This randomness helps the ants adapt to changes in the environment, as a few ants spontaneously seek out more direct ways to their food source.
In biology, self-organized systems can be found at a range of scales, from the patterns of proteins inside cells to the socially complex colonies of honeybees that collectively build nests and forage for nectar.
Randomness in sunflower self-organization
So, could random, irregular circumnutations underpin the sunflowers’ self-organization?
My colleagues and I set out to explore this question by following the growth of young sunflowers we planted in the lab. Using cameras that imaged the plants every five minutes, we tracked the movement of the plants to see their circumnutatory paths.
We saw some loops and spirals, and lots of jagged movements. These ultimately appeared largely random, much like Darwin’s carnation. But when we placed the plants together in rows, they began to move away from one another, forming the same zigzag configurations that we’d seen in the previous study.
We analyzed the plants’ circumnutations and found that at any given time, the direction of the plant’s motion appeared completely independent of how it was moving about half an hour earlier. If you measured a plant’s motion once every 30 minutes, it would appear to be moving in a completely random way.
We also measured how much the plant’s leaves grew over the course of two weeks. By putting all of these results together, we sketched a picture of how a plant moved and grew on its own. This information allowed us to computationally model a sunflower and simulate how it behaves over the course of its growth.
A sunflower model
We modeled each plant simply as a circular crown on a stem, with the crown expanding according to the growth rate we measured experimentally. The simulated plant moved in a completely random way, taking a “step” every half hour.
We created the model sunflowers with circumnutations of lower or higher intensity by tweaking the step sizes. At one end of the spectrum, sunflowers were much more likely to take tiny steps than big ones, leading to slow, minimal movement on average. At the other end were sunflowers that are equally as likely to take large steps as small steps, resulting in highly irregular movement. The real sunflowers we observed in our experiment were somewhere in the middle.
We wanted our model sunflowers to do the same thing. So, we made it so that two plants that get too close to each other’s shade begin to lean away in opposite directions.
Finally, we wanted to see whether we could replicate the zigzag pattern we’d observed with the real sunflowers in our model.
First, we set the model sunflowers to make small circumnutations. Their shade avoidance responses pushed them away from each other, but that wasn’t enough to produce the zigzag—the model plants stayed stuck in a line. In physics, we would call this a “frustrated” system.
Then, we set the plants to make large circumnutations. The plants started moving in random patterns that often brought the plants closer together rather than farther apart. Again, no zigzag pattern like we’d seen in the field.
But when we set the model plants to make moderately large movements, similar to our experimental measurements, the plants could self-organize into a zigzag pattern that gave each sunflower optimal exposure to light.
So, we showed that these random, irregular movements helped the plants explore their surroundings to find desirable arrangements that benefited their growth.
Plants are much more dynamic than people give them credit for. By taking the time to follow them, scientists and farmers can unlock their secrets and use plants’ movement to their advantage.
Citation:
Sunflowers make small moves to maximize sun exposure—physicists can model them to predict how they grow (2024, September 16)
retrieved 16 September 2024
from https://phys.org/news/2024-09-sunflowers-small-maximize-sun-exposure.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.
The new wave of artificial intelligence—so-called AI—is bringing with it promises as well as threats.
By assisting workers, it can raise productivity and boost real wages. By making use of large, underutilized data, it can improve outcomes in services including retailing, health and education.
The risks include deepfakes, privacy abuse, unappealable algorithmic decisions, intellectual property infringement and wholesale job losses.
Both the risks and the potential benefits seem to grow by the day. On Thursday, Open AI released new models it said could reason, performing complex calculations and drawing conclusions.
Separately, the government is consulting about mandatory guardrails for AI in high-risk settings, which would function as a sort of checklist for what developers should consider alongside a voluntary safety standard.
Here’s my thinking: most of the potential uses of AI are already covered by existing rules and regulations designed to do things such as protect consumers, protect privacy and outlaw discrimination.
These laws are far from perfect, but where they are not perfect the best approach is to fix or extend them rather than introduce special extra rules for AI.
AI can certainly raise challenges for the laws we have—for example, by making it easier to mislead consumers or to apply algorithms that help businesses to collude on prices.
But the key point is that laws to control these things exist, as do the regulators experienced in enforcing them.
The best approach is to make existing rules work
One of Australia’s great advantages is the strength and expertise of its regulators, among them the Competition and Consumer Commission, the Communications and Media Authority, the Australian Information Commissioner, the Australian Securities and Investments Commission, and the Australian Energy Regulator.
Their job ought to be to show where AI is covered by the existing rules, to evaluate the ways in which AI might fall foul of those rules, and to run test cases that make the applicability of the rules clear.
It is an approach that will help build trust in AI, as consumers see they are already protected, as well as provide clarity for businesses.
AI might be new, but the established consensus about what is and is not acceptable behavior hasn’t much changed.
Some rules will need to be tweaked
In some situations, existing regulations will need to be amended or extended to ensure behaviors facilitated by AI are covered. Approval processes for vehicles, machinery and medical equipment are among those that will increasingly need to take account of AI.
And in some cases, new regulations will be needed. But this should be where we end up, not where we begin. Trying to regulate AI because it is AI will, at best, be ineffective. At worst, it will stifle the development of socially desirable uses of AI.
Many uses of AI will create little if any risk. Where potential harm exists, it will need to be weighed against the potential benefits of the use. The risks and benefits ought to be judged against real-world, human-based alternatives, which are themselves far from risk-free.
New regulations will only be needed where existing regulations—even when clarified, amended or extended—are inadequate.
Where they are needed, they ought to be technology-neutral wherever possible. Rules written for specific technologies are likely to quickly become obsolete.
Last mover advantage
Finally, there’s a lot to be said for becoming an international “regulation taker.” Other jurisdictions such as the European Union are leading the way in designing AI-specific regulations.
Product developers worldwide, including those in Australia, will need to meet those new rules if they want to access the EU and those other big markets.
If Australia developed its own idiosyncratic AI-specific rules, developers might ignore our relatively small market and go elsewhere.
This means that, in those limited situations where AI-specific regulation is needed, the starting point ought to be the overseas rules that already exist.
There’s an advantage in being a late or last mover. This doesn’t mean Australia shouldn’t be in the forefront of developing international standards. It merely means it should help design those standards with other countries in international forums rather than striking out on its own.
The landscape is still developing. Our aim ought to be to give ourselves the best chance of maximizing the gains from AI while providing safety nets to protect ourselves from adverse consequences. Our existing rules, rather than new AI-specific ones, ought to be where we start.
Citation:
Opinion: The best way to regulate AI might be not to specifically regulate AI (2024, September 16)
retrieved 16 September 2024
from https://techxplore.com/news/2024-09-opinion-ai-specifically.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.
A new study allays fears that rerouting flights to avoid forming climate-warming contrails could result in inadvertently making climate warming worse.
Researchers from Sorbonne University and the University of Reading found that for most flights that form contrails in the North Atlantic, the climate benefit of avoiding the contrail outweighs the extra carbon dioxide emitted from flying a different route.
Contrail avoidance requires comparing the climate impacts of carbon dioxide and contrails, called CO2 equivalence. Different methods have been proposed, and the choice of which has been largely political. Scientists feared that some choices could be misleading, making avoidance seem beneficial for the climate when it is in fact damaging.
The study, published 15 September in Atmospheric Chemistry and Physics, finds that for a large majority of North Atlantic flights, contrail avoidance would benefit climate regardless of the choice of CO2 equivalence.
Contrails explained
Contrails—the white lines left behind planes in the sky—can trap heat in the atmosphere and contribute to global warming.
The new study builds on previous research that suggested planes could be rerouted to avoid contrail formation, potentially reducing climate impact. However, the benefits of avoiding contrails against the drawbacks of extra CO2 emissions were unclear.
Prof Nicolas Bellouin, co-author at the University of Reading, said, “Rerouting flights to avoid contrails could in theory reduce the climate impact of aviation and make air travel more sustainable. Our findings lift a major obstacle against implementing contrail avoidance, but we now need better forecasting and real-world trials to make this work in practice.”
The new findings show that regardless of how the trade-off between contrail avoidance and increased CO2 emissions is measured, rerouting rarely worsens climate effects unintentionally. The study looked at nearly half a million flights over the North Atlantic in 2019 to estimate how much warming was caused by the carbon dioxide emissions from these flights and any contrails they formed.
The researchers first examined how current flight routes would warm the world over time. They estimate that the CO2 emissions and contrails from these flights will have warmed the climate by about 17 microKelvins (μK) in 2039, 20 years later, and 14 μK in 2119, 100 years later. A microKelvin is a very tiny unit of temperature change.
Then the researchers imagined a situation where planes could avoid all contrails by using just 1% more fuel. In this case, the total warming would decrease significantly. By 2039, warming would be reduced by about 5 μK, which is 29% less than without rerouting. By 2119, it would be about 2 μK (14%) less.
The researchers used nine different ways to measure climate impact. In most cases, all these methods agreed that rerouting flights would be good for the climate, as long as the planes successfully avoid contrails as predicted.
The researchers emphasize that there is still much uncertainty in predicting exactly where contrails will form and how much warming they cause. They suggest focusing initial rerouting efforts on flights that form the most warming contrails, where the climate benefit is clearest.
More information:
Audran Borella et al, The importance of an informed choice of CO2-equivalence metrics for contrail avoidance, Atmospheric Chemistry and Physics (2024). DOI: 10.5194/acp-24-9401-2024
Citation:
Contrail avoidance is less likely to damage climate by mistake than previously thought, researchers find (2024, September 16)
retrieved 16 September 2024
from https://phys.org/news/2024-09-contrail-climate-previously-thought.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.
New research highlights the precarious status of one of Madagascar’s most threatened amphibians, the harlequin mantella (Mantella cowanii), revealing small population sizes and the possible extirpation of the species from several of its historic habitats.
The findings underscore the urgent need for targeted conservation action to prevent the species from slipping further towards extinction. The findings are published in the journal PeerJ.
A dire situation
Amphibians around the world are facing unprecedented population declines, and Mantella cowanii is no exception. The study, which focused on confirming the frog’s presence at historical localities and estimating its population size and survival rates, paints a concerning picture.
Out of the 13 known localities for M. cowanii, the frog was detected in only eight, with three populations potentially extirpated. However, the research also uncovered two new populations, offering a glimmer of hope amidst the otherwise bleak outlook.
Repeated annual surveys at three sites revealed population sizes ranging from as few as 13 to 137 adults, with one site experiencing an alarming 80% reduction in population size from 2015 to 2023. The study also discovered that M. cowanii has a relatively slow life history pace, with adult survival rates between 0.529 and 0.618, and a maximum lifespan in the wild reaching nine years or more. This slower life history makes the species particularly vulnerable to extinction.
The primary threats to Mantella cowanii include illegal collection for the international pet trade and ongoing habitat degradation. These pressures are compounded by the frog’s limited distribution and small, isolated populations, which further increase its extinction risk.
Conservation recommendations
In light of these findings, the researchers recommend immediate and sustained conservation efforts. Key actions include:
Continued population monitoring: Ongoing monitoring of M. cowanii populations is critical to track trends, measure the effectiveness of conservation actions, and detect declines before they become irreversible.
Reassessing the IUCN Red List Status: Given the recent population estimates and trends, there is a strong case for reassessing the species’ status on the IUCN Red List. Mantella cowanii was last assessed in 2014 as Endangered, but it may now qualify for the more severe Critically Endangered status.
Engaging Local Communities: Conservation efforts should involve local communities in monitoring programs, which can foster pride and support for the species while adding value to conservation initiatives.
Further Research: Additional research is needed to better understand the relative impact of threats such as disease, illegal trade, and habitat loss. This knowledge is essential to developing effective conservation strategies.
Searching for Mantella cowanii in Madagascar. Credit: Dr. Devin Edmonds
Global implications
The challenges faced by Mantella cowanii are emblematic of a broader global trend in amphibian population declines. As amphibians are often indicators of environmental health, the decline of M. cowanii signals broader ecological challenges that could have far-reaching impacts.
The time to act is now. Conservationists, researchers, and policymakers must come together to implement the recommended actions and ensure the survival of Mantella cowanii. Without decisive intervention, we risk losing this unique species forever.
More information:
Devin Edmonds et al, Small population size and possible extirpation of the threatened Malagasy poison frog Mantella cowanii, PeerJ (2024). DOI: 10.7717/peerj.17947
Citation:
Urgent conservation efforts needed: Possible extirpation of the threatened Malagasy poison frog Mantella cowanii (2024, September 16)
retrieved 16 September 2024
from https://phys.org/news/2024-09-urgent-efforts-extirpation-threatened-malagasy.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.
