Remote collaboration software tools, such as Zoom or Google Docs, have become essential for teamwork. But they often overlook the fact that people do not all approach collaboration in the same way, according to researchers at North Carolina State University (NCSU).
Scientists report that they have now developed a new human-computer interaction (HCI) method called RemoteCollabEval (RCE) to identify barriers to collaboration and inclusivity, allowing designers and developers to build software features that better support diverse teamwork styles.
The work is part of the broader HCI field, which examines how people use digital systems and how interfaces can be optimized for clarity and ease of use.
“At present, most remote collaboration platforms are evaluated by designers and developers using established HCI inspection methods,” says Sandeep Kuttal, PhD, the principal investigator behind the work and an associate professor of computer science at NCSU. “One of the most widely used inspection methods is a ‘groupware walkthrough,’ where designers essentially play out how a collaborative effort might unfold between two or three hypothetical users. However, these methods typically assume all users behave in similar ways.”
As senior author of a paper, “Equity by Design: A New HCI Method for Surfacing Inclusivity Issues in Remote Collaboration Software,” that will be presented at the ACM Designing Interactive Systems Conference (DIS 2026) in Singapore, from June 13-17, Kuttal notes that “It’s well-established that people from various backgrounds often have different collaboration and communication styles. “Existing HCI inspection methods don’t account for these differences, which limits how inclusive and effective these tools can be. That’s what we set out to address.”
Six key personality facets
As a first step, the researchers drew on established social science and software engineering research to identify six key personality facets that influence collaborative behavior:
Leadership style: Does the individual take a democratic or authoritative approach?
Interruption style: Does the individual interrupt others or wait for cues?
Non-verbal cues: Is the individual expressive or reserved in digital spaces?
Relationship-seeking: Does the individual focus on building rapport or primarily on achieving goals?
Social awareness: Is the individual attentive to or unaware of what their teammates are doing?
Collaborative self-efficacy: How confident is the individual in the group’s ability to perform?
The researchers then created hypothetical users called “personas,” which are detailed representations of different types of users that incorporate descriptions of each of the six facets. These personas allow designers to simulate interpersonal friction and uncover “inclusivity bugs” that might otherwise go unnoticed during standard testing.
“Because we have descriptions of all six facets for each persona, we can incorporate those key characteristics into our assessment of how well a given platform allows for effective collaboration between people of different backgrounds,” explains Kuttal.
The team then modified existing groupware walkthrough methods, requiring designers and developers to explicitly consider these six facets as part of the process and created a specialized walkthrough. This combination of personas that account for personality facets and the specialized walkthrough forms the RCE method.
As a proof-of-concept study, the scientists recruited 29 undergraduate and graduate students and split them into 10 teams. Five teams inspected an existing remote collaboration platform using the conventional Groupware Walkthrough method; the other five teams inspected the same platform using RCE.
“The teams who used the RCE method identified six times more inclusivity issues than the conventional method,” continues Kuttal. “Essentially, RCE did a better job of identifying when conflicting styles would make collaboration between personas difficult. This is important, because identifying these challenges gives designers and developers an opportunity to modify features and user interfaces to improve these remote collaborative platforms. And, ultimately, to improve collaboration itself.
“Because RCE is a standardized, systematic method, it can be used by designers and developers anywhere. It doesn’t require a huge budget, or an expensive research effort. It’s a method that can easily be used to make these platforms better.”
Cytokinetics said Tuesday that its drug Myqorzo significantly improved heart failure symptoms and cardiovascular fitness in patients with non-obstructive hypertrophic cardiomyopathy, an inherited heart disorder.
The results, reported in a company press release, achieved the dual efficacy goals of a Phase 3 clinical trial, called ACACIA, with statistical significance.
Cytokinetics is in the early days of Myqorzo’s commercial launch as a treatment for the more severe “obstructive” form of hypertrophic cardiomyopathy, or HCM. The successful outcome of the ACACIA study, if also cleared by regulators, could greatly expand the number of HCM patients eligible for treatment — and boost the drug’s peak sales to $5 billion annually, according to analyst forecasts.
Every few centuries, changes in how information moves reshape how societies govern themselves. The printing press spread vernacular literacy, helping give rise to the Reformation and, eventually, representative government. The telegraph made it possible to administer vast nations like the US, accelerating the growth of the modern bureaucratic state. Broadcast media created shared national audiences, which in turn fueled mass democracy.
We are now in the early stages of another such shift. Faster than many realize, AI is becoming the primary interface through which we form beliefs and participate in democratic self-governance. If left unchecked, this shift could further strain America’s already fragile institutions. But it could also help address long-standing problems, like lagging civic engagement and deepening polarization. What happens next depends on design choices that are already being made, whether we know it or not.
Start with what might be called the epistemic layer—how we come to know things. People are increasingly relying on AI to know what is true, what is happening, and whom to trust. Search is already substantially AI-mediated. The next generation of AI assistants will synthesize information, frame it, and present it with authority. For a growing number of people, asking an AI will become the default way to form views on a candidate, a policy, or a public figure. Whoever controls what these models say therefore has increasing influence over what people believe.
Technology has always shaped the way citizens interact with information. But a new problem will soon arise in the form of personal AI agents, which can change not only how people receive information but how they act on it. These systems will conduct research, draft communications, highlight causes, and lobby on a user’s behalf. They will inform decisions such as how to vote on a ballot measure, which organizations are worth supporting, or how to respond to a government notice. They will, in a meaningful sense, begin to mediate the relationship between individuals and the institutions that govern them.
We’ve already seen with social media what happens when algorithms optimize for engagement over understanding. Platforms do not need to have an explicit political agenda to produce polarization and radicalization. An agent that knows your preferences and your anxieties—one shaped to keep you engaged—poses the same risks. And in this case the risks may be even more difficult to detect, because an agent presents itself as your advocate. It speaks for you, acts on your behalf, and may earn trust precisely through that intimacy.
Now zoom out to the collective. AI agents and humans could soon participate in the same forums, where it may be impossible to tell them apart. Even if every individual AI agent were well-designed and aligned with its user’s interests, the interactions of millions of agents could produce outcomes that no individual wanted or chose. For example, research shows that agents displaying no individual bias can still generate collective biases at scale. And setting aside what agents do to each other, there is what they do for their users. A public sphere in which everyone has a personalized agent attuned to their existing views is not, in aggregate, a public sphere at all. It is a collection of private worlds, each internally coherent but collectively inhospitable to the kind of shared deliberation that democracy requires.
Taken together, these three transformations—in how we know, how we act, and how we engage in collective governance—amount to a fundamental change in the texture of citizenship. In the near future, people will form their political views through AI filters, exercise their civic agency through AI agents, and participate in institutions and public discussions that are themselves shaped by the interactions of millions of such agents.
Today’s democracy is not ready for this. Our institutions were designed for a world in which power was exercised visibly, information traveled slowly enough to be contested, and reality felt more shared, if imperfectly. All of this was already fraying long before generative AI arrived. And yet this need not be a story of decline. Avoiding that outcome requires us to design for something better.
On the informational layer, AI companies must ramp up existing efforts to ensure that models’ outputs are truthful. They should also explore some promising early findings that AI models can help reduce polarization. A recent field evaluation of AI-generated fact checks on X found that people with a variety of political viewpoints deemed AI-written notes more helpful than human-written ones. The paper is yet to be peer-reviewed, but that is a potentially revolutionary finding: AI-assisted fact-checking may be able to achieve the kind of cross-partisan credibility that has eluded most manual human efforts. Greater understanding of and transparency about how models make these assertions and prioritize sources in the process could help build further public trust.
On the agentic layer, we need ways to evaluate whether AI agents faithfully represent their users. An agent must never have an agenda of its own or misrepresent its user’s views—a technically daunting requirement in domains where users may have not explicitly stated any preferences. But faithful representation also cannot become an accessory to motivated reasoning. An agent that refuses to present uncomfortable information, that shields its user from ever questioning prior beliefs or fails to adjust to a change of heart, is not acting in the person’s best interest.
Finally, on the institutional level, policymakers should hurry to harness AI’s potential to make governance more responsive and legitimate. Several states and localities are already using AI-mediated platforms to conduct democratic deliberation at scale, building on research showing that AI mediators can help citizens find common ground. As agents become increasingly common participants in public input processes—and there is already evidence that bots are skewing those processes—identity verification for both humans and their agentic proxies must be built in from the start.
What is needed is a new generation of democratic infrastructure, technological and institutional, built for the world that is actually here. Failing to design for democratic outcomes, in a domain this consequential, means designing for something else. And the history of unaccountable power does not leave much room for optimism about what that something else tends to be.
Andrew Sorota and Josh Hendler lead work on AI and democracy at the Office of Eric Schmidt.
About 10 years after a breast surgeon we interviewed returned to Dubai to practice, a colleague stopped her in a hospital corridor to tell her: “It’s great — since you came back I no longer see those advanced cases of breast cancer.”
She had not stopped to notice. “You get on the wheel,” she told us in an interview for our research. When you return to your home country after training abroad, you run your clinics, train your nurses, drive out to women’s groups in the hinterlands to give talks, design educational videos that explain breast self-examination without showing a breast, because that is what the censors will pass. You argue to have the word “breast” printed on your medical license; the authorities prefer “chest.” You establish a support group, then a drop-in center, likely the only cancer drop-in center in the Middle East. Somewhere in the middle of all that, you alter what late-stage breast cancer looks like in your country.
When synthetic biologists sketch gene circuits, they usually think in terms of promoters, repressors, and transcription factors—biochemical parts that toggle genes on or off. But DNA is not a flat schematic. It’s a physical polymer that twists, coils, and buckles as genes are transcribed. A pair of new papers from MIT and collaborators shows that this physicality could suggest approaches to controlling the output of gene circuits.
In a recent Science study titled “Gene syntax defines supercoiling-mediated transcriptional feedback,” researchers demonstrate that the order and orientation of neighboring genes—what they call gene syntax—can reshape local DNA supercoiling and, in turn, amplify or suppress the expression of adjacent genes.
“Syntax will be really useful for dynamic circuits. Now we have the ability to select not only the biochemistry of circuits, but also the physical design to support dynamics,” said Katie Galloway, PhD, an assistant professor of chemical engineering at MIT.
The team engineered human cell lines and hiPSCs with synthetic two‑gene reporter circuits arranged in tandem, divergent, or convergent configurations. Their earlier modeling predicted that divergent syntax should boost the expression of both genes, while tandem syntax should suppress the downstream gene. “The thing that we were trying to solve in this paper was: When you put two genes on the same piece of DNA, how does their physical interaction become coupled?” said Galloway. The experimental results matched those predictions: divergent circuits amplified both genes, while tandem circuits showed strong upstream‑to‑downstream repression, with effects reaching up to 25‑fold.
To understand why, the researchers used Region Capture Micro‑C, a high‑resolution genome‑folding mapping technique, to visualize how transcription reshapes DNA. When a gene was activated, the DNA downstream tightened into plectonemes—twisted structures that hinder RNA polymerase binding—while upstream DNA loosened. “Supercoiling impacts transcription of adjacent genes by altering RNA polymerase binding, forming a feedback loop,” the authors of the first paper wrote.
The second paper, published in Nature Biomedical Engineering and titled “STRAIGHT-IN Dual: a platform for dual single-copy integrations of DNA payloads and gene circuits into human induced pluripotent stem cells,” introduced STRAIGHT‑IN Dual, a platform that enables simultaneous, allele‑specific, single‑copy integration of two DNA constructs into hiPSCs. This system allowed the team to “investigate how promoter choice and gene syntax influence transgene silencing and how these design features affect reporter expression and forward programming of hiPSCs into neurons, motor neurons, and endothelial cells,” according to the authors of the second paper.
Using STRAIGHT‑IN Dual, the researchers also demonstrated a practical application: a divergent circuit expressing two components of a yellow fever antibody produced higher output than other configurations.
“This is really exciting because we can coordinate gene expression in ways that just weren’t possible before,” Galloway said. “Now that we understand the syntax, I think this will pave the way for us to program dynamic behaviors.
“If you want coordinated expression, a divergent circuit is great. If you want something that’s either/or, you can imagine using a convergent or tandem circuit, so when one turns on, the other turns off, and you can alternate pulses,” Galloway added.
Scientists headed by a team at St. Jude Children’s Research Hospital have found that in individuals with the life-threatening blood disorder aplastic anemia (AA), different blood stem cells within the same person independently acquire gene mutations that allow cells to escape the immune attack. Through their study, the team, together with collaborating institutions, used state-of-the-art genomic techniques to profile 619 children and adults with AA. The study showed that for some patients, these “rescuing” stem cell clones were enough to restore blood production and provide long-term remission.
“We found that each patient with aplastic anemia that escapes autoimmunity has multiple, independent genetic events in different blood stem cells that allow those cells to escape autoimmunity,” said Marcin Wlodarski, MD, PhD, St. Jude Department of Hematology. “Stem cells silence the risk HLA allele through several mechanisms, and our data show that these events are protective, benign events that don’t cause progression to MDS or leukemia, even when the rescued clones grow and dominate the bone marrow.”
Corresponding author Wlodarski and colleagues reported on the study, which they say includes the largest pediatric cohort of its kind reported to date, in Nature Genetics. In their paper titled “High-resolution single-cell mapping of clonal hematopoiesis and structural variation in aplastic anemia,” the team wrote, “These findings reveal parallel evolutionary pathways used by hematopoietic cells to evade immune attack.”
Aplastic anemia is a rare, life-threatening bone marrow failure (BMF) syndrome where patients are unable to make enough blood cells due to the immune system’s attack on hematopoietic stem and progenitor cells (HSPCs). The condition can progress to myelodysplastic syndrome (MDS) and leukemia.
In AA, autoreactive T cells target and destroy blood stem cells that display peptides on a specific protein on their surface. These are encoded by the human leukocyte antigen (HLA) gene. Each person inherits one copy of this gene from each parent, which can have different variations. People with aplastic anemia often carry a particular “risk” HLA allele (gene variant) that is thought to trigger the disease. As the authors noted, “While the precise mechanism underlying HSPC recognition by autoimmune T cells remains elusive, specific human leukocyte antigen (HLA) alleles are overrepresented in patients with AA compared with healthy controls, suggesting a role in aberrant immune recognition.”
Some blood stem cells evade the immune attack by acquiring changes that silence the risk HLA allele. This can happen via loss-of-function HLA mutations or through uniparental isodisomy 6p (UPD6p), where the risk allele is replaced with a non-risk allele. “HLA loss, manifesting as uniparental disomy of chromosome 6p (UPD6p) or loss-of-function (LOF) mutations in HLA, is postulated to inactivate HLA risk alleles (presumed to mediate autoantigen presentation), effectively shielding HSPCs from autoimmune attack,” the investigators noted. Two other types of escape in blood stem cells are known: paroxysmal nocturnal hemoglobinuria (PNH) or mutations in clonal hematopoiesis (CHIP) genes. However, it was unclear if all these changes arise in a single stem cell or arise independently to help the blood stem cells hide from the immune system. It was also unclear how this process of immune evasion impacted clinical outcomes and cancer risk.
“The clinical implications of clonal alterations in AA vary,” the investigators stated. “HLA loss is generally considered a nonmalignant adaptive lesion, large PNH clones require complement inhibitor therapy, and CHIP-mutant clones may be associated with MDS, thereby necessitating hematopoietic stem cell transplantation (HSCT).”
(L to R) Corresponding author Marcin Wlodarski, MD, PhD, and lab member Diantha Van De Vlekkert, both of the St. Jude Department of Hematology, and second author Sushree Sahoo, PhD, formerly of the St. Jude Department of Hematology. [St. Jude Children’s Research Hospital]
Blood stem cells give rise to all other blood cells, meaning their progeny are genetically identical, including any mutations gained over time. The relative abundance of a specific stem cell’s genetic “clones” measures the genetic diversity of these blood-making cells. Using single-cell analyses, the researchers showed that protective mutations happen independently in different blood stem cells and not sequentially within a single cell. These independent clones then repopulate the marrow without being found and killed by the immune system. “We saw that patients with blood stem cell clones that escape autoimmunity can improve their blood counts,” Wlodarski said. “We also learned that these clones do not indicate an increased risk for leukemia. On the contrary, they often indicate the possibility of long-lasting remission.”
To assess these clones, the scientists analyzed bone marrow and blood samples from 619 (256 children and 363 adults) patients with AA. “We present a high-resolution genomic landscape in AA patients using single-cell targeted DNA/protein sequencing, PacBio long-read whole-genome sequencing (WGS), and single-cell WGS,” they explained. They found that overall, 69% of patients carried at least one acquired change: HLA mutations or UPD6p clones were found in 16%, PNH clones in 44%, and CHIP mutations in 21%.
First author Masanori Yoshida, MD, PhD, St. Jude Department of Hematology, then established and applied a single-cell DNA sequencing assay to simultaneously profile mutations and cell-surface proteins of 304,902 single cells from 48 samples. The study was complemented by long-read whole-genome sequencing and single-cell whole-genome sequencing.
The experiments showed that acquired mutations are just as common in children as in adults, but in pediatric patients, 65% of the CHIP mutations occurred in just three genes (BCOR, BCORL1, and ASXL1), compared with 27% in adults. Because age-related CHIP mutations are not expected to preexist in children, these mutations seem to be immune-escape events acquired in response to the autoimmune attack. “In children, where preexisting CHIP is not expected, mutations in these three genes may represent bona fide immune escape mechanisms arising in direct response to T-cell-mediated attack,” the authors stated.
To understand how these protective events arise and to count them precisely, the authors performed whole-genome sequencing on many single blood stem cells. They expected to see one to three events per individual; instead, they found a median of three per patient, and in one patient, 15 independent clones, all resulting in the loss of the risk HLA allele, showing convergent evolution to escape a strong immune attack. “Strikingly, HLA loss clones emerge independently through mutational events that converge on inactivating a single specific HLA risk allele, with up to 15 clones per patient identified using the scWGS platform … Our analyses reveal that somatic alterations in AA arise as independent clones rather than through sequential acquisition, and most patients carry multiple independent clones,” the investigators noted.
That extreme diversity pointed to an unusual, convergent evolutionary process, so the scientists reconstructed a phylogenetic “family tree” of individual blood stem cells by reading all mutations acquired throughout life in single whole genomes. This method enabled them to pinpoint each clone’s origin. “We had expected that these mutations occur right before disease onset,” Wlodarski said. “But we found some of these HLA-loss clones arose many years before clinical diagnosis.”
The team also showed that long-lived, rescued clones had higher expression of CD34, a surface marker for blood stem and progenitor cells. This suggests that CD34 enrichment could serve as a biomarker of long-lasting recovery. In addition, clones with loss of HLA risk alleles and CHIP mutations almost never co-occurred in the same cells, indicating that HLA loss provides enough of a proliferative advantage on its own that additional CHIP mutations, which can predispose to MDS, are not selected. So, they appear to act as protective events against their MDS and leukemia evolution.
“Clones with higher CD34+ expression levels measured in our scDNAseq/protein analysis, particularly those with HLA-loss alterations, demonstrated long-term fitness, multilineage contribution, and were often associated with stable blood counts and prolonged treatment-free intervals,” the team pointed out. These results challenge prior assumptions about when and how protective clones arise in aplastic anemia, and their presence can be a factor in restoring blood formation.
“Aplastic anemia shows us convergent evolution in miniature: Multiple independent mutational events arise in different cells, all leading to the same escape from autoimmunity,” Wlodarski said. “It shows the amazing nature of human hematopoiesis to cure itself from bad actors, like the autoimmune T cells, and reconstitute the bone marrow.” In their paper, the team concluded, “These findings enhance our understanding of clonal dynamics in AA and provide a foundation for future precision medicine approaches to address BMF in this life-threatening syndrome.”
<strong>Background:</strong> Stroke remains a primary cause of long-term disability worldwide, with upper-limb deficits affecting up to 80% of survivors acutely and 40% chronically. These deficits lead to considerable effects on their independence and overall quality of life. Conventional rehabilitation therapies are most effective when initiated shortly after a stroke, yet many patients face barriers to ongoing therapy post discharge. Recent advancements in low-cost rehabilitation systems, particularly those using virtual reality (VR) technologies, offer promising alternatives for enhancing upper-limb recovery. <strong>Objective:</strong> Given the burden on health care systems and the limitations in access to high-intensity postdischarge rehabilitation, this study aimed to evaluate the feasibility, acceptability, and usability of an upper-limb adaptive mirror therapy using VR and myoelectric control for the rehabilitation of patients with chronic stroke developed through a user-centered design approach. <strong>Methods:</strong> In this study, a total of 12 community-dwelling survivors of chronic stroke (mean age 52.9, SD 16.0 years; 4 female) with moderate to severe upper-limb impairments were enrolled. Participants were stratified by age (young: 18-55 years; older: 56-80 years) and impairment level (Fugl-Meyer Assessment-Upper Extremity score: 18-36=severe; 37-54=moderate). Acceptability was assessed for each session by patient self-evaluation of satisfaction and motivation through a visual analog scale, while the therapist assessed the patient’s participation in therapy using the Pittsburgh Participation Rehabilitation Scale. Usability was measured with the User Satisfaction Evaluation Questionnaire scale and feasibility through the NASA (National Aeronautics and Space Administration) Task Load Index cognitive workload indices. <strong>Results:</strong> Patients reported a significant increase in satisfaction from the intermediate to the final assessment (T1: 72% vs T2: 85%; <i>P</i>=.01) and stable high motivation levels. Differences in participation and motivation were observed based on impairment levels, with no effect of age. Usability ratings remained high (>80%) throughout the intervention, with no significant differences between baseline and end line (<i>P</i>=.56). Cognitive workload assessments showed a significant reduction over time, in perceived cognitive (<i>P</i>=.04), performance (<i>P</i>=.007), and effort demands (<i>P</i><.001). Impairment level significantly influenced perceived workload, with participants with more severe impairment reporting higher cognitive, physical, temporal, and effort demands (all <i>P</i><.001), while age did not contribute to variability in acceptability, usability, or workload measures. <strong>Conclusions:</strong> VR therapy was found to be feasible, under adaptive task conditions, ensuring stable performance across patients. The protocol was usable and acceptable among patients with chronic stroke, especially those with moderate impairment, supporting its potential as a user-centered digital rehabilitation tool, warranting further investigation in controlled and home-based settings. <strong>Trial Registration:</strong> ClinicalTrials.gov NCT07103122; https://clinicaltrials.gov/study/NCT07103122
Background: Evidence-based psychological interventions are usually not accessed by marginalized groups such as refugees. Culturally adapted psychological interventions have reported larger effect sizes than nonadapted psychological interventions. However, the cultural adaptation of interventions is a lengthy process, entailing a challenge. One potential solution to overcome this challenge is the use of artificial intelligence (AI). Objective: The aim of this study was to investigate and compare the perceived cultural relevance and acceptability of 2 common cognitive behavioral therapy (CBT) techniques when translated and culturally adapted by AI versus a human psychologist. Methods: In a 2×2 factorial design, the text generator type (AI vs human psychologist) and the CBT technique (cognitive restructuring vs behavior modification) were compared. CBT technique texts translated and culturally adapted either by AI or by a human psychologist were blindly rated using the Cultural Relevance Questionnaire and the Theoretical Framework of Acceptability. Raters were Arabic-speaking refugees and immigrants, aged between 18 and 69 years, residing in Sweden, Denmark, and Germany. Raters were randomly allocated to 1 of 4 conditions. Each condition consisted of 2 stimuli. Two-factor between-subject design analyses were used to analyze the data. Results: A significant main effect of the text generator domain type (=.02; η²=0.045) was found in the first rating, with texts adapted by the AI domain perceived as more culturally relevant than those adapted by the human domain. No significant main effect of the CBT technique was found in the first rating (=.10; η²=0.022). There were no differences in the second rating. Regarding acceptability, no significant main effects of text generator domain type (=.09; η²=0.024) or the CBT technique (=.88; η²=0.001) were found in either of the ratings. Conclusions: CBT technique materials adapted by AI may be perceived as similarly culturally relevant as those adapted by a human psychologist. This finding implies the potential to accelerate the cultural adaptation of psychological interventions. However, AI still needs to be used with caution and in accordance with rigorous safety standards and robust frameworks.
<img src="https://jmir-production.s3.us-east-2.amazonaws.com/thumbs/717971a3020c51a80c4b35fb1e059041" />
![(L to R) Corresponding author Marcin Wlodarski, MD, PhD, and lab member Diantha Van De Vlekkert, MSc, both of the St. Jude Department of Hematology, and second author Sushree Sahoo, PhD, formerly of the St. Jude Department of Hematology. [St. Jude Children's Research Hospital]](https://www.genengnews.com/wp-content/uploads/2026/05/Low-Res_10560975884-022_JPG-300x200.jpeg)
