Resignations Rise Due to Forest Carbon Accounting Rule Disagreements about Threats to Scientific Integrity

     Carbon offsetting is way up due to AI data center developers, ie., tech companies, seeking to offset their growing carbon emissions. Debates about accounting for forest carbon storage are pitting scientists against the forest products industry that wants less stringency on their credits.

     Tim Searchinger, the Princeton University scholar serving as the World Resources Institute‘s technical director for land-related issues, recently resigned his position in protest. University of Pennsylvania senior fellow Danny Cullenward stepped down from the Greenhouse Gas Protocol’s Independent Standards Board in June over the same dispute. Searchinger told Bloomberg that he resigned:

“…because developments with the Protocol undermined that integrity in both the content of the rules and the process behind them, and because he was unable to correct them.”

     According to Wood Central, the debate is about how much to attribute to human activity vs. natural activity.

“The Greenhouse Gas Protocol, a standard-setter overseen in part by the World Resources Institute, is drafting its first guidance on forest carbon accounting, and the fight centres on two competing approaches to attributing the carbon forests absorb, store, and release. Activity-based accounting separates changes caused by people from those driven by natural processes, while the rival managed land proxy treats every change on managed forest land as human-caused.”

     Opponents say there is double counting when natural carbon removal is counted and credited as human-caused removal.

“Bloomberg found most scientists and non-governmental organisations either backed activity-based accounting or criticised the proxy, with a smaller band of scientists writing in favour alongside industry professionals.”

     The issue is whether or not this is greenwashing, or more accurately, how much greenwashing will be accepted.

     Microsoft is one of the biggest buyers of carbon offsets and one of the biggest buyers of forest carbon offsets.

“Microsoft, one of the world’s biggest buyers of wood-linked carbon removal, has already agreed to buy more than 4.8 million tonnes of credits from improved forest management across American forestland in five states. Surging AI emissions are testing the tech giant’s pledge to become carbon negative by 2030.”

     An article about this in The Cool Down states that the problem is simply greenwashing, whether a company can claim sustainability branding without really adding much real carbon storage benefit. They explain the debate between activity-based accounting and managed land proxy as follows:

“Activity-based accounting tries to sort out which carbon changes come from human actions and which stem from natural processes. Managed land proxy, by contrast, counts every carbon change on managed forest land as human-caused.”

     Choosing the managed land proxy method means that corporate disclosures will be distorted and less scientifically accurate.  

     The Greenhouse Gas Protocol standards board has yet to make a decision and will accept comments through February 2027, before making a decision. At present, either accounting method may be used as long as the method used is disclosed.

     There is also the possibility of using a hybrid model, the “managed land proxy plus” model, which is safeguarded against being exploited irresponsibly. However, several scientists believe that it is not enough and there will still be inaccuracies in the real amount of carbon stored by the corporate interests, akin to greenwashing.

“Vaughan Andrews, Weyerhaeuser's senior program manager for sustainability, helped develop a revised version he called "managed land proxy plus" that includes safeguards so it is "applied responsibly." It's a "solution that can be implemented at scale," Melissa Gallant, The Nature Conservancy's senior climate adviser, told Bloomberg.”

“Others are not persuaded. Jennifer Skene, global forest policy director at the Natural Resources Defence Council, argued that a tool designed for national climate inventories was never intended for corporate accounting.”

"You don't want any actor who has a climate goal to be reporting removals from the atmosphere without acknowledging or somehow taking into account the fact that some of those forest-based removals really have nothing to do with their corporate action," Nathan Truitt, an American Forest Foundation executive vice president, said. "That is absolutely not what will happen if we adopt [managed land proxy plus]."

    

References:

 

Princeton expert resigns over forest carbon rules, citing threat to 'scientific integrity'. Leslie Sattler. The Cool Down, August 22, 2026. Princeton expert resigns over forest carbon rules, citing threat to 'scientific integrity'

Top Scientist Quits as Forest Carbon Rules Face Greenwashing Claims. Wood Central. August 17, 2026. Forest Carbon Accounting Fight Sees Top Scientist Quit WRI | Wood Central

 

Textile Emissions, Pollution, and Waste: Fast Fashion Accelerates It: Fashion Industry Challenges, Data, Regulations, Sustainability, and Recycling Efforts

 

     The fashion industry, which relies on textile manufacturing and dyeing, is an outsized contributor to climate change, air pollution, and water pollution. It is responsible for about 10% of global carbon emissions. Thus, efforts towards sustainability have been a focus for the industry for years. Marketing had reflected this trend until the recent affordability challenges. Elsa Ohlen, in a CNBC article, notes:

“…despite the marketing blitz, fashion executives admit that most consumers, battered by a persistent cost-of-living crisis, aren't willing to pay more for an ethically produced product, or "sustainability premium." Instead, shoppers have become increasingly selective and price-sensitive, looking for value above all else.”

     This, she says, has created a retail paradox, where more expensive options are being promoted, but only the cheaper options are selling. The disruption in the Middle East has led to higher oil and oil products prices as well as higher prices for petroleum-based synthetic fibers like polyester. The cost crunch leaves less in the budget for environmental considerations. This is another example where profitability enables environmental spending and lack of profitability disables it.   

     Sustainability and ESG have long been important in the fashion industry, where unfair labor practices and animal welfare issues have been exposed in the past. Pollution and carbon emissions add considerably to the industry’s vulnerability to scrutiny. Scandals can taint brand equity.

     Some fashion companies want to move toward the use of recycled materials, but say they need incentives.  

"There are no short-term incentives to use circular materials, on the contrary, you have to pay more," said Helmersson. "We're acting in a market where there's no price on nature," she said, adding that the industry needs increased taxation on nature – such as taxing the resource footprint of products and services – and decreased taxation on labor.

     I disagree with that carbon tax approach since it would increase costs for consumers, but government incentives could help, maybe something like a tax credit for using recycled materials. The article notes that textile-to-textile recycling is in need of an infrastructure buildout.

     According to a report by the Boston Consulting Group:

“About 80% of discarded clothing ended up in landfills or incinerators, 12% was reused, and less than 1% was recycled into new fibers, the report found.”

     The EU is preparing to regulate textile waste with one goal of disincentivizing overproduction, which has been implicated in the trends toward “fast fashion,” where lower-quality, more or less disposable clothing is mass-produced to be inexpensive. The EU’s proposed rule, designed to address waste and greenwashing, is the "Ecodesign for Sustainable Product Regulation" (ESPR). Failure to comply will lead to financial penalties and reputational damage.

     According to Science Insights:

“Textile waste is any discarded material that comes from clothing, fabrics, or fiber-based products. This includes worn-out garments, factory scraps from manufacturing, unsold inventory, and household items like bedding, towels, and curtains.”

     They note two broad categories of textile waste: pre-consumer and post-consumer waste. Pre-consumer textile waste is the waste from manufacturing, including fabric offcuts, rejected materials, and surplus stock that does not sell. Post-consumer waste is what we consumers discard. Fast fashion’s shorter lifespan means more is needed to replace it.

     The composition of textile waste determines how long it will last in the environment while in use, but especially afterward. Studies have shown that about 70% of textile waste is made of natural fibers, mainly cotton and linen. Synthetic fibers like polyester, nylon, and acrylic make up a growing proportion of waste. These do not readily break down in landfills. Fabrics that blend natural and synthetic fibers are troublesome as waste because it is difficult to separate those fibers in recycling steps. Synthetic fibers can take up to 200 years to fully decompose in a landfill. Those synthetic fibers can also leach toxic chemicals that join the landfill leachate. Metal-based additives are in common use in the industry. According to Science Insights, these include zinc compounds used as heat stabilizers and in dyes, copper added as complex dyes, antimony-containing flame retardants, and lead and cadmium salts used as stabilizers in synthetic polymers.

     They also note how the synthetic fibers break down into microplastics :

“Tiny fiber fragments between 1 micrometer and 5 millimeters long are released from synthetic garments during washing, machine drying, and even regular wearing. These microplastic fibers wash into waterways through household drains, passing through many wastewater treatment systems and reaching rivers, lakes, and oceans.”

“Once in the water, these fibers are ingested by aquatic animals, disrupting their metabolic processes. From there, they enter the human food chain through seafood, sea salt, and drinking water. Any manufacturing process that weakens fiber structure, including sanding, brushing, and bleaching, increases the volume of fragments shed during later use. A single load of laundry can release hundreds of thousands of these fibers, and the cumulative effect across billions of households is enormous.”

     Most textile waste is incinerated or landfilled. Fiber-to-fiber recycling, where the fibers of one garment are extracted and remade into another garment, remains uncommon. They list the reasons why below:

     Most textile recycling is “downcycling,” where the discarded clothing is made into lower-quality goods like insulation, cleaning rags, or stuffing material. This delays but does not eliminate the waste stream.

     The EU requires extended producer responsibility (EPR), where the manufacturer is responsible for tracking and collecting post-consumer discarded textile waste. The producers are required to cover the costs of collecting, sorting, and recycling. Some discarded clothing of reasonable quality is resold domestically or exported to secondhand markets in Africa, Asia, and Latin America. Around 30% of textile waste considered for recycling ends up in a landfill. Without collection, it all ends up in a landfill.

     The Boston Consulting Group’s analysis quantified textile waste, where it goes, and other industry statistics. 

     They note the loss in potential raw materials:

“Each year, textile waste worth an estimated $150 billion in raw materials value is lost—resources that are extracted, processed, and then quickly discarded. Recovering even a quarter of these wasted resources could offset the combined annual materials expenditures of the world’s 30 largest fashion companies. The environmental and economic case is clear: reducing textile waste would conserve valuable resources and minimize environmental and social impact.”

     92% of the industry’s greenhouse gas emissions come from producing textiles. Incineration, landfilling, and open dumps account for the rest.

     The table below explores barriers to textile recycling in terms of waste generated, sorting, recycling, and processing.

     Potential solutions include designing garments to be recycled, building out collection infrastructure and increasing consumer awareness about collection, developing automated sorting systems, and investing more in waste management. Below they propose five ways to scale up a circular economy for textile waste: 1) Promote demand for textiles with recycled fibers, 2) Collect more waste, 3) Modernize sorting systems, 4) Scale effective recycling solutions, and 5) Invest in innovation.

     They think that success is dependent on industry-wide collaboration or central orchestration by the public sector.

“The textile industry can draw inspiration from other sectors that have successfully transformed through a combination of public subsidies, regulatory incentives, targeted investments, and consumer engagement.”

“The path toward a circular textile economy is in sight, but it requires decisive action across the entire value chain. Industry players must boost demand for recycled fibers, streamline textile waste collection, upgrade sorting infrastructure, expand proven recycling methods, and invest in cutting-edge recycling technologies.”

     The recent sale of millennial fashion darling Everlane to fast-fashion giant Shein does not seem to bode well for sustainable fashion. Shein has had issues with working conditions and environmental impacts. However, Everlane’s CEO says it will remain an independent company and continue its focus on producing sustainable clothing. Everlane epitomized the direct-to-consumer business model. Everlane had taken on too much debt and had been seeking a sale. Shein has been accused of buying the company as a form of greenwashing its own fast fashion products or at least diluting its reputation for unsustainable fashion.

 

 

References:

 

A strange paradox has taken hold of the global fashion industry. Elsa Ohlen. CNBC. May 20, 2026. A strange paradox has taken hold of the global fashion industry

Shein acquiring Everlane feels like the end of the millennial sustainable fashion dream. Juliana Kaplan. Business Insider. May 22, 2026. Shein acquiring Everlane feels like the end of the millennial sustainable fashion dream

What Is Textile Waste? Definition, Sources & Impact. Science Insights. March 14, 2026. What Is Textile Waste? Definition, Sources & Impact - ScienceInsights

Spinning Textile Waste into Value. Rohan Sajdeh, Catharina Martinez-Pardo, Alexander Meyer zum Felde, Tom Lange, Eleonora Tieri, and Elian Evans. Boston Consulting Group. August 12, 2025. Tackling the Global Crisis of Textile Waste | BCG

 

 

Empathy and Politics: ‘Suicidal Empathy,’ the Extents of Moral Concern, the Oppressor/Victim Mentality, and Parochialism vs. Universalism

   

     According to Merriam-Webster, the meaning of empathy is:

“The action of understanding, being aware of, being sensitive to, and vicariously experiencing the feelings, thoughts, and experience of another.”

     Another similar definition of empathy is:

“The ability to understand and share the feelings of another.”

     Caring and concern for others and their feelings and thoughts seem important and like common sense to many of us humans.  

     Elon Musk famously or infamously said that empathy is the fundamental weakness of Western civilization. I don’t think that is correct at all. Certainly, empathy can be misplaced, and that happens frequently where people think they are doing the most good when the data show they are not. It is also difficult to know and to prioritize different actions that benefit others. Net harm and net benefit should be considered. However, immediate relief from suffering should be prioritized in most cases, in my view.

 

Empathy and Compassion Are Similar but Not the Same

     I have come across several people who say they are empaths, that they feel the emotions of others. Some of those do not like what they see as their “gift” and would prefer not to have it. To me, that suggests that empathy is not the same as compassion. Buddhists define compassion more or less as the wish to reduce the suffering of others, themselves, and all beings. The definition of empathy suggests that it is an experience, often one which the person does not choose. Compassion, on the other hand, is most often seen as a choice to act to help others, or in a lesser form, a sincere wish to help others. Empathy in itself is not good or bad, and in order to better understand the plight of others directly, it can be very valuable in fostering the choice to be compassionate. Ideally, it is like a useful precursor to compassion, but that is not always the case. Compassion does not require empathy but can be aided by it. We can be compassionate toward all beings, including the worst of beings, without sharing their feelings or seeing through their eyes.

 

The Left is Not Always More Empathetic

     While the political left is known for empathy more than the political right, those on the right tend to give a little more to charitable causes, which perhaps reflects slightly greater wealth, but also a significant amount of concern that can be construed as empathy. One case where empathy seemed to be missing among those on the far left is the October 7 massacre of civilians in Israel, where many of those on the far left said things like Israel brought this on themselves or that it was bound to happen due to Israeli policies. That was quite shocking to me and many others: that one of the most bloody and brazen attacks on civilians that included cold-blooded murder in many forms, rape, and long-term torture, including psychological torture, was met with such attitudes.

     Those further to the left often espouse an oppressor/victim mentality. This idea derives from Karl Marx and informs their views about conflicts around the world. It has become a leading narrative among them and has informed ideas like critical race theory, intersectionalism, and postcolonial theory, and is part of what has become known as post-Marxism. I do not believe that is the best way to evaluate such conflicts, whether wars or class conflicts. It may provide a lot of fodder for talking points, but it is basically a bunch of hooey.   

 

Suicidal Empathy

     An article in Vox by aid Jilani notes that:

“Suicidal empathy,” the phrase coined by Canadian marketing professor and conservative influencer Gad Saad, is suddenly on the lips of everyone right-of-center”

     Musk promoted Saad’s book, “Suicidal Empathy.” Musk suggested that liberals have empathy for the criminals but not for the victims of crimes, due to their habit of sticking up for those at the bottom. Conservatives who buy into the suicidal empathy narrative point to a 2019 study by a group of psychologists that concluded those on the right have greater concern for those closer to them, including close family and friends, while those on the right extended their concern more to those further from them, including foreigners, homeless people, and animals. Conservatives used the study to conclude that liberals value the lives of distant foreigners and strangers over those of their own family and community. However, a new paper in Political Psychology rejects that view, noting that a greater outward extent of concern does not diminish concern for those nearer. Thus, it is not a zero-sum situation where there is only so much empathy to go around. I think this is a more correct view. The quote below is a snippet from the abstract, which follows the quote.

“In unconstrained settings, greater concern for distant entities does not predict reduced concern for close others; ideological differences reflect how far moral concern extends outward, not compromised ingroup prioritization.”

     The abstract of the 2019 paper, below, presents the study as a comparison of parochialism and universalism. Parochialism refers to expressing more compassion for more structured and less encompassing entities, and universalism refers to more compassion for less structured and more encompassing entities. However, I don’t think being conservative has to mean restricting one’s circle of moral concern, nor do I think being liberal has to mean extending it. There is another Buddhist idea known as equanimity, or impartiality, where one extends compassion while not distinguishing between those near and far. Compassion is extended to all beings without exception. This is easy to say but perhaps much harder to do since we probably have instincts to protect those closer to us. However, those instincts may be motivated by selfish concerns.

     Jilani notes that Kyle Fiore Law, an assistant psychology professor at the University of Utah and one of the authors of the 2026 study, thinks that the 2019 study was not robust enough to give the full prioritizations of liberals and conservatives.

“The 2019 studies measured how far moral concern extends, and they did that well,” Law wrote to me in an email. “They were not designed to detect whether concern for distant others competes with concern for close others, because both tasks built an outward-expanding structure (e.g., from close humans to distant humans to non-human life to eventually the natural world and non-sentient targets) into the measure itself.”    

     The research suggests simply that liberals tend to expand their moral concern further outward than conservatives. This makes sense. Liberals don’t value strangers’ lives more than close friends’ lives, and it is a bit absurd to think that they do. They simply tend to have more concern about those strangers than conservatives tend to have.

“In our data, extending concern outward did not come with reduced concern for close others,” he said.

     Jilani writes:

“Tradeoffs are a fact of life. Whether we should be willing to modestly increase the national debt or taxes to increase foreign aid, for instance, is a question of values rather than something that can be empirically derived.”

     The recent shutdown of USAID is a shining example of a shrinking moral circle of concern by conservatives. Thus, I would say the problem is not liberals with wide circles of concern but conservatives with smaller ones.

     Jilani writes:

“As a species, we succeeded not just because of our capacity to defend our in-groups — our family, friends, community, nations — but because of our ability to understand the interests and values of those far away.”

 

Idiot Compassion

     Buddhists are usually quite familiar with the notion of idiot compassion, where one may harm another more by seeking to ease their suffering in ways that don’t work. It often occurs when one wants to ease their own selfish concerns, rather than truly help others.

     According to a Microsoft Copilot search:

“The term idiot compassion was coined by Tibetan Buddhist teacher Chögyam Trungpa and further explained by his student Pema Chödrön. It refers to being kind or accommodating in ways that ultimately cause more harm than good. This occurs when compassion is expressed without wisdom, boundaries, or courage, often to avoid conflict or discomfort. Essentially, it is compassion that serves the giver’s desire to feel good rather than the recipient’s actual needs.”

     I am not sure if this is the best example, but countries giving money and resources to Gaza, with much of it ending up going to the terrorist group Hamas, which has effectively ruled the region for decades. They used that money to build tunnels and to build up their military capabilities instead of making their society prosperous. It may not be a good example because some of that money did and does relieve suffering. However, some of it is siphoned off by a group that does not care about the suffering of its enemies nor of its citizens. For instance, the UN paid salaries for many employees who were essentially operatives of the terrorist organization.

References:

 

The right says “suicidal empathy” is destroying the West. Here’s the truth. Zaid Jilani. Vox. July 28, 2026. The right says “suicidal empathy” is destroying the West. Here’s the truth.

Ideological differences in the expanse of the moral circle. Adam Waytz, Ravi Iyer, Liane Young, Jonathan Haidt & Jesse Graham. Nature Communications. volume 10, Article number: 4389 (2019). Ideological differences in the expanse of the moral circle | Nature Communications

Ideological differences in moral concern reflect circle expansion, not inversion. Kyle Fiore Law, Seoyeon Bae, Liane Young, and Stylianos Syropoulos. Political Psychology. Volume 47. Issue 3. June 2026. Ideological differences in moral concern reflect circle expansion, not inversion - Law - 2026 - Political Psychology - Wiley Online Library

Idiot Compassion and Mindfulness. Derek Beres. Big Think. October 30, 2013. Idiot Compassion and Mindfulness - Big Think

 

Biosolids Management Costs Set to Rise for Wastewater Utilities Over the Coming Decade, According to Market Report

    According to a market report by Bluefield Research, U.S. municipal wastewater utilities are facing growing costs for biosolids management. They estimate that US$36.4 billion will be needed over the next decade, nearly double current costs. Biosolids management is being affected by the closing of older incinerators, landfill capacity constraints, and PFAS contamination management. These all make biosolids management more expensive.

“According to a new Insight Report, U.S. Municipal Biosolids Management: Drivers, Trends, and Forecasts, 2025–2035, total U.S. municipal biosolids management spending will surge, growing from US$2.5 billion in 2025 to more than US$4.8 billion annually by 2035.”

     Cities, towns, and wastewater utilities will bear the costs.

     According to Smart Water Magazine:

“…more than 23,900 treatment plants across the country process wastewater effluent for approximately 267 million people. In doing so, they generate about 6.3 million dry metric tons of biosolids each year—a nutrient-rich byproduct that has long been applied to farmland, converted into fertilizers, incinerated, or sent to landfills.”

     Incinerators are closing due to stricter air quality rules. Community opposition and permitting obstacles are slowing landfill expansions. Concerns about concentrated PFAS contamination in biosolids are also slowing land application for fertilizer use.

“Many municipal wastewater utilities, particularly in urban areas, are finding that their preferred low-cost disposal methods are increasingly becoming less available or just simply more expensive,” says Pat Byrne, an analyst at Bluefield Research. “As a result, utilities are investing in more affordable methods and technologies like drying and dewatering, while also giving greater consideration of advanced destruction systems.”

     Bluefield forecasts that the Northeast will account for US$13.6 billion, more than one-third of total U.S. municipal biosolids spending over the next decade. Landfill tipping fees have been increasing in the Southeast amid continued population growth. This will drive spending to US$9.8 billion, or 27% of the national total. Land application remains more accessible across the Midwest and West. However, logistical challenges and growing local opposition may change that. The graphic below shows that landfilling will remain the most common method. It also shows regional differences in strategies expected over the next decade.

“There’s no one-size-fits-all approach,” Byrne notes. “Utilities are making decisions based on local capacity, alternative costs, and community sentiment—all key factors in driving regional differences in how biosolids are managed across the country.”

     Some states are banning land application of biosolids. Maine enacted the first ban in 2022. New monitoring rules and contaminant thresholds are also being developed. The EPA’s recent designation of PFOS and PFOA as hazardous substances also presents liabilities for those who choose land application.

“Rising biosolids hauling costs and limited local disposal capacity are pushing utilities to invest in dewatering and drying technologies aimed at reducing biosolids volume and moisture before disposal.”

     Other innovations include advanced destruction and thermal processing systems.

“There’s a growing opportunity for companies that can help utilities adapt to these shifting market dynamics,” Byrne explains. “Whether through advanced destruction technologies or improved operational models, utilities are looking for partners that can deliver sustainable, cost-effective solutions for biosolids management.”

  

References:

 

U.S. utilities face $4.8 billion biosolids challenge amid rising costs. Bluefield Research. Smart Water Magazine. 5 December 2025. U.S. utilities face $4.8 billion biosolids challenge amid rising costs – Smart Water Magazine

‘Enhancing Soil Science Research with Multi-Agent Artificial Intelligence Systems’: Summary and Review of Paper in ‘Frontiers in Science’

      Having explored some multi-agent AI workflows in subsurface analysis, I have some ground for understanding the basis for using multiple AI agents to optimize insights from data.

     Machine Learning (ML), a subset of AI, has been utilized in soil science to solve specific problems and has proved useful in a predictive capacity. What ML-based soil science can’t do is robust investigations of the underlying cause-and-effect mechanisms resulting in the generation of hypotheses. The authors cite a study by Khanifar that evaluated LLMs for soil science knowledge and “found that, on average, they could correctly answer only up to 65% of questions from advanced soil science examinations.” That is OK but certainly not good enough.

     The authors note that utilizing multi-agent AI in soil science should emulate successes in other fields:

“An AI agent is a software system that uses AI to interact with its information environment, gather information, and perform tasks to achieve specific goals.”

“The vision of AI agents transforming scientific discovery is gaining momentum in fields such as chemistry, materials science, and biomedicine, where complex systems, vast datasets, and multidisciplinary knowledge must be integrated.”

     The goals of the study are listed below:

     Ideally, an AI agent becomes a co-scientist, a partner. AI agents may have the ability to solve complex problems in complex domains like soil science. The figure below shows how five AI agents with specific tasks can be integrated in order to solve the problem of assessing the effects of climatic changes on soil organic matter retention. 

     Soil carbon storage is affected by soil heat, which increases microbial activity that breaks down organic matter faster, and by soil moisture.

     Multi-agent AI has emerged as a very effective means of advancing scientific research and can accelerate discoveries.

“A multi-agent system can act as an orchestrator, selecting, configuring, and executing these tools within a coordinated workflow, moving beyond the current fragmented use of individual technologies. The agents possess perceived reasoning and planning capabilities that extend well beyond simple data processing.”

“These systems emulate collaborative scientific workflows involving brainstorming, critique, consensus-building, and iterative refinement.”

     The table below summarizes the key features of multi-agent AI models in scientific research. There are four models: 1) general multi-agent AI systems, 2) AI co-scientists (which can refine hypotheses), 3) astro agents (which can analyze large data sets, finding clusters and utilizing pattern recognition), and 4) spatial agents used in biomedical research.

     These models share core principles and capabilities:

     Below, the authors propose a multi-agentic system for soil analysis to be utilized for hypothesis generation and experiment design. It integrates key components from leading multi-agent AI models. Agents and sub-agents are given the goal of proceeding from knowledge acquisition to hypothesis generation to evaluation and feedback, and to arrive at a pool of hypotheses. A Meta-Review Agent then aggregates these evaluation results and ranks the hypotheses, which are then selected and evaluated by actual scientists.

“The next leap forward lies in intelligent automation: the development of multi-agent AI systems that not only process and predict but also have the technical abilities to emulate perception, reasoning, planning, and collaboration. These systems have the potential to work alongside human experts, navigating complex scientific questions, integrating diverse data sources, generating hypotheses, and designing adaptive experiments. Embracing this paradigm would allow soil science to harness AI’s full potential—not to replace researchers but to empower them in addressing the global challenges of soil security, climate resilience, and sustainable land use.”    

     The authors give four types of multi-agent systems they believe can transform soil science research.

1)        Integrated soil digital twin agents – the digital twin can use algorithms to detect and identify variables that affect nutrient depletion, water stress, compaction, or erosion hotspots, for instance. The system could model land management options or soil system responses to climate extremes. It could generate reports and maps.

2)        Soil microbiome analysis agents – this would require metagenomics or meta-transcriptomics to evaluate microbial species and be integrated with detailed environmental parameters such as soil chemistry, mineralogy, moisture, temperature, or vegetation types. Soil-microbial and plant-microbial interactions can be unraveled. It could evaluate the effects of soil amendments or herbicides. It could help in understanding the functions of the many proteins produced by soil microbes. For instance, it could evaluate and test nutrient availability under drought conditions. Continent-wide microbial distributions can be determined and used to evaluate soil-plant-climate interactions.

3)        Climate change impact and adaptation agents – a system such as this could integrate climate models with soil process models. Organic carbon stocks and nutrient and moisture availability could be evaluated. It could evaluate cropping systems or water conservation techniques. It could evaluate the soil’s ability to sequester carbon. It could design soil sampling strategies and optimize fertilization rates. These multi-agent systems can handle complexity and automate workflows. They warn here that AI cannot replace scientists, especially when conditions arise that are outside the specificities of their training.

4)        Hypothesis generation and experimental design agents – here they note that multi-agentic systems can automate literature analysis and synthesis and “identify knowledge gaps and propose novel, testable hypotheses about soil processes.” They could also design experiments to test those hypotheses.

“Agents should be able to communicate their reasoning, present findings clearly, and solicit and incorporate expert feedback.”

          The authors next present a case study of hypothesis generation in soil carbon saturation research. It is mainly a feasibility study for hypothesis generation. Knowing soil organic carbon (SOC) saturation is important in determining the capacity of a soil to retain and sequester carbon and for assessing soil health and fertility. They explain below the debates about carbon saturation and why testable hypotheses are needed to answer relevant questions.  

“A central concept of organic carbon persistence is soil carbon saturation, particularly regarding mineral-associated organic carbon (MAOC). This concept suggests that mineral surfaces, especially within clay and silt fractions, have a finite capacity to form organo-mineral complexes due to limited reactive surface area. However, its operational relevance is under debate. Empirical studies by Begill et al., Heinemann et al., and Poeplau et al. challenge the idea of a clear saturation plateau, showing near-linear increases in MAOC even under high carbon inputs, implying that biological and environmental constraints may often limit SOC accrual before theoretical saturation is reached. Conversely, Cotrufo et al. maintain the physical validity of the saturation concept, arguing that observed inconsistencies stem from methodological artefacts, such as the misclassification of particulate organic matter (POC) as MAOC, and emphasize the need to identify real-world factors that prevent soils from reaching their saturation potential.”

     They utilized Manus AI as their multitask agent. It has Planner, Execution, and Verification agents that work collaboratively. The AI system was prompted to scan recent scientific literature both in and out of the field of soil science to generate contrasting hypotheses, simulate peer review, and rank them in terms of novelty, feasibility, plausibility, and scalability.

“In summary, the simulated multi-agent reviews found the hypotheses to be clear, coherent, and diverse, empirically varied but scientifically valuable, with several requiring advanced methods for testing, and well-differentiated with no major redundancy, supporting a broad and balanced research agenda.”

     The hypotheses that are more difficult to test and those with less scalability across soil types tended to be ranked lower.

“Table 5 summarizes the conceptual understanding of soil MAOC saturation based on Georgiou et al. (58), comparing it with the AI-generated hypotheses.”

     Below is a section of the summary:

“AI systems can help researchers gather information and rapidly generate hypotheses that are grounded in empirical evidence and theoretical frameworks. This could reduce the time and effort typically required to transition from exploratory reading to focused scientific inquiry.”

“Equally important is the enhanced rigor introduced through simulated peer review. This needs to be further improved through a step-by-step sequence where researchers evaluate the basis of the hypotheses. This step requires human input and interaction. For example, soil scientists stated that they would like to see a more fundamental understanding of interactions between clay minerals and OM, as most studies use only clay content to determine or evaluate carbon stock. However, it remains uncertain whether the hypotheses generated by AI are entirely novel or simply well-synthesized from existing knowledge. The “black box” nature of the AI system also does not allow for probing of where the knowledge was derived from.”

     In a section about challenges and future directions, the authors note that data availability, scarcity, fragmentation, and heterogeneity due to variable sources, as well as lack of standardization have been identified as challenges.

“Developing and implementing explainable AI (XAI) methods tailored to soil science applications is essential. Current XAI for LLMs and generative models mainly provide transparency and provenance rather than explanations of internal reasoning. While this can improve trust and usability, it does not offer causal insight into decision processes, reinforcing the need for human oversight and uncertainty-aware deployment.”

“Predictive uncertainty in multi-agent workflows remains insufficiently explored. Sensitivity to prompts, stochastic planning, variable evidence quality, and measurement errors can propagate through agent chains, increasing the risk of spurious correlations in a data-poor and methodologically heterogeneous domain such as soil science.”

     They also point to difficulties in integrating AI systems with physical processes such as automated sample collection, managing sensor networks, and integrating with lab equipment. Computational costs are another concern.

     The quotes below explore the human-AI interface amid AI’s current constraints.

“Agents must be more than pattern recognizers; they need a degree of scientific understanding.”

“Creativity as an essential component of the (human) scientific process should also be nurtured, especially since this remains a limitation of multi-agent AI systems. AI models are fundamentally dependent on the data they are trained on. While these agents can autonomously search for and retrieve literature, their effectiveness is shaped by how they access, filter, and prioritize sources. This data-dependency constrains their ability to think “outside the box”, a quality often crucial for scientific breakthroughs. Although AI excels at recombining and synthesizing existing ideas, its capacity for true creativity and the generation of genuinely novel concepts remains uncertain. One potential strength, however, lies in AI’s ability to explore knowledge across disciplines that are rarely considered by soil scientists, thereby expanding the conceptual search space. The central question remains: can AI genuinely create or does it merely reconstruct and reconfigure what is already known?”

     The quotes below are from the conclusion and continue the assessment of AI capabilities.

“These intelligent collaborators can dramatically accelerate the pace of discovery, transforming vast datasets into meaningful insights. By exploring multiple solution pathways and building on existing knowledge, AI can expand the boundaries of scientific exploration while still relying on human judgment to interpret, refine, and apply the findings.”

“While AI can emulate aspects of expert reasoning, it cannot replace the contextual judgment, creativity, and critical interpretation that human scientists bring to the research process. As such, AI should be viewed as an augmentative tool, enhancing but not substituting human-led scientific inquiry.”

 

References:

 

Enhancing soil science research with multi-agent artificial intelligence systems. Budiman Minasny. Alex McBratney, José A.M. Demattê, Mercedes Román Dobarco, and Pete Smith. Frontiers in Science. Vol 4. May 20, 2026. Frontiers | Enhancing soil science research with multi-agent artificial intelligence systems