Tag: News Approved

Posted on

The Download: your stake in OpenAI, and the Treasury’s AI warning

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.

Your family’s $300 stake in OpenAI

Sam Altman’s proposal that Americans should share in the wealth created by AI is back in the spotlight, with reports that he is discussing giving the US government a 5% stake in OpenAI. At the company’s current valuation, that stake would be worth roughly $320 per American household.

The idea is meant to address concerns that AI companies are benefiting from human-generated work without compensating creators, while also easing fears that AI will cause a collapse of the labor market by providing a safety net. 

The details, however, remain unclear. Indeed, the offer may be more powerful as a political narrative than as a policy plan.

Read the full story on what the dividend proposal reveals about the future of AI.

—James O’Donnell

This article is from The Algorithm, our weekly AI newsletter. Sign up to receive it in your inbox every Monday.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

 1 A leaked Treasury report compares the AI market to the dotcom bubble
 Which contradicts the administration’s public optimism about AI. (NOTUS)
+ Fears that the market is overinflated are growing. (Reuters $)
+ And AI profits are hiding bigger risks in earnings reports. (FT $)
+ What even is the AI bubble? (MIT Technology Review)

2 Samsung profits have jumped 1,800% on booming AI chip sales
It just reported its third consecutive record quarterly profit. (BBC)
+ But its shares slumped over fears that the AI boom will stall. (Reuters $)
+ That boom has turned Samsung into a $1 trillion company. (CNBC)
 
3 A US cyber agency is using Mythos to audit government code
Sources say CISA is tapping Anthropic’s model to search for bugs. (Reuters $)
+ Agencies are using it despite Anthropic’s feud with the White House. (Axios)
 
4 Illinois’ governor has signed the nation’s strongest frontier AI law
It’s designed to protect citizens from AI risks. (Gizmodo)
+ US lawmakers are clashing over AI rules. (MIT Technology Review)
 
5 A hidden tracker in Claude Code has been exposed and removed
It secretly monitored users in China. (WP $)
+ Critics said it shows Anthropic’s willingness to surveil users. (Ars Technica)
+ The company has also found a hidden “thinking” space in Claude. (Axios)

6 Russia is suspected of flying drones over Europe from a shadow fleet
The flights were reportedly launched from commercial ships. (Ars Technica)
+ Europe has a drone-filled vision for future wars. (MIT Technology Review)
 
7 A controversial AI “actor” is set to star in its first feature film
Tilly Norwood will debut in a comedy-drama called “Misaligned.” (Variety)
+ A major actors union has lambasted the AI creation. (NBC News)

8 AI costs are driving US companies toward Chinese models
Businesses are hunting for cheaper model alternatives. (CNBC)
+ Chinese AI labs are betting big on open source. (MIT Technology Review)
 
9 Researchers have shown quantum proofs can beat classical ones
They found a problem that classical proofs can’t solve. (Quanta)

10 Earth will never be swallowed by the sun, according to new models
But it probably won’t be much fun to live here by that point anyway! (Wired $)

Quote of the day

“The goal might be to make machines in our image. But what I fear is that—perhaps without even quite noticing—we remake ourselves in theirs.” 

—Reporter Sarah O’Connor sounds a note of caution in her new book, We Are Not Machines, the Guardian reports.

One More Thing

mammoth walking on a strand of DNA

KATE DEHLER


Adventures in the genetic time machine

Eske Willerslev, a specialist in recovering DNA from old bones and objects, has made numerous breakthroughs. These include recovering the first more or less complete genome of an ancient human and 2.4-million-year-old genetic material from Greenland, revealing that today’s Arctic desert was once a forest with poplar, birch, and mastodons.

These findings are part of a wave of discoveries from what’s being called an “ancient-DNA revolution.” 

Beyond revealing stories of human migration and vanished ecosystems, scientists believe ancient DNA can unearth clues about modern diseases. It could even lead to a better food supply for our warming world. “And can we get that?” Willerslev asks. “Yes, I believe we can.”

Discover how ancient DNA could rescue the future. 

—Antonio Regalado

We can still have nice things

A place for comfort, fun, and distraction to brighten up your day. (Got any ideas? Drop me a line.)

+ Underworld’s electric set from EDC Las Vegas 2026 has been released as a full concert video.
+ This photographic journey through global soccer culture captures the mad passion of fandom around the world.
+ Feeling unhinged? Me too. This playlist of gloriously intense classical music sympathetically captures the mood.
+ If you’re looking for visual inspiration, this collection of graphic design archives from across the web is a goldmine.

Posted on

Prototypical graph based deep label propagation with semantic augmentation for cross-subject and cross-session EEG emotion recognition

Electroencephalogram (EEG)-based emotion recognition faces significant generalization challenges in cross-subject and cross-session settings, primarily due to the inherent non-stationarity, individual differences, and semantic deficiency of EEG signals. To address these challenges and enhance the universality of affective brain-computer interface systems, this study proposes a novel unsupervised domain adaptation framework named Prototypical Graph-based Deep Label Propagation with Semantic Augmentation (PGDLP). PGDLP seamlessly integrates three core components—prototypical semantic augmentation, prototype-graph deep label propagation, and prototypical alignment—into an end-to-end optimized system. Specifically, class-wise multivariate normal distributions are constructed using source-domain feature statistics to augment target-domain features semantically, bridging the domain gap and mitigating semantic insufficiency. An adaptive similarity graph based on prototype-semantic distances is designed to optimize pseudo-label quality while reducing computational complexity. It is combined with linear projection and an exponential moving average (EMA) for dynamic refinement. Dual intra- and inter-domain alignment losses with an adaptive balance factor are introduced to enhance intra-class compactness and inter-domain transferability, thereby facilitating the learning of discriminative and domain-invariant features. Extensive experiments are conducted on three benchmark datasets (SEED, SEED-IV, DEAP) under four rigorous evaluation protocols (cross-subject cross-session, cross-subject single-session, within-subject cross-session, cross-database). Overall, PGDLP achieves superior recognition accuracy and generalization performance compared with most state-of-the-art methods across the majority of evaluation protocols. PGDLP also presents strong robustness against label noise, stable hyperparameter performance, and fast convergence. The results demonstrated that PGDLP outperforms state-of-the-art methods in recognition accuracy and generalization, with verified robustness to label noise, stable hyperparameters, and efficient convergence. This study provides a promising solution for unsupervised cross-domain EEG emotion recognition and offers valuable insights for domain adaptation research on other physiological signals.
Posted on

Development of a transformation model to analyze horizontal saccadic velocity using electrooculography: a pilot feasibility study

IntroductionSaccadic eye movements are established biomarkers in neuroscience and clinical neurology, with video-oculography (VOG) serving as the gold standard for measurement. However, the high cost, bulky equipment, and poor portability of VOG systems restrict their clinical utility. Electrooculography (EOG) provides a practical alternative, but quantitative conversion of EOG-derived measurements into VOG-equivalent values remains insufficiently established. This study aimed to develop and validate a mathematically derived transformation model for estimating VOG-equivalent horizontal saccadic velocities from EOG recordings.MethodsFour healthy adults underwent simultaneous EOG and VOG recordings while performing controlled horizontal gaze shifts. Based on a current-source model of the corneal potential, an analytical relationship between EOG voltage velocity and angular eye velocity was derived. Multiple high-pass filter settings were systematically evaluated to identify optimal signal-processing conditions. Transformation equations were derived from the pooled horizontal-saccade dataset and further evaluated using leave-one-subject-out (LOSO) analysis.ResultsThe theoretical model predicted a linear relationship between EOG- and VOG-derived saccadic velocities. Among the tested filter settings, a 0.3 Hz high-pass combined with a 35 Hz low-pass filter yielded the best overall agreement. Under this condition, the final transformation model produced a common slope coefficient of 0.146 °/μV for both movement directions, with an additional direction-specific intercept of −82.37 °/s for rightward saccades. Converted EOG-derived velocities showed no significant differences from measured VOG-derived velocities. LOSO validation demonstrated stable transformation coefficients (mean slope = 0.147 °/μV, mean intercept = −82.47 °/s) and maintained agreement across individuals.ConclusionA biophysically derived and experimentally validated EOG-to-VOG transformation model can provide accurate estimates of horizontal saccadic velocity under appropriate filtering conditions. These findings support the feasibility of quantitative saccadic analysis using EOG, providing a practical alternative to VOG.
Posted on

Lower limb rehabilitation in chronic incomplete SCI: a randomized controlled trial protocol comparing combined TMS-tSCS neuromodulation vs. tSCS alone

BackgroundSpinal cord injury (SCI) affects 15.4 million people worldwide, with a substantial proportion of incomplete SCI patients remaining non-ambulatory, highlighting the importance of motor function recovery and mobility in rehabilitation. While transcutaneous spinal cord stimulation (tSCS) has emerged as a promising non-invasive neuromodulation technique for enhancing motor recovery, the therapeutic potential of combining tSCS with transcranial magnetic stimulation (TMS) remains largely unexplored. This combination may leverage the complementary mechanisms of supraspinal and spinal neuromodulation to enhance corticospinal tract plasticity and functional motor outcomes.ObjectiveTo evaluate the efficacy and safety of combined TMS-tSCS intervention compared to tSCS alone for improving lower extremity motor function in individuals with chronic incomplete spinal cord injury.MethodsThis prospective, randomized, controlled, assessor-blinded clinical trial will enroll 60 participants with chronic (>12 months post-injury) incomplete spinal cord injury (AIS C or D) aged 18–65 years from Alexandra Hospital, Singapore. Participants will be randomized 1:1 to receive either combined TMS-tSCS (intervention group) or tSCS with sham TMS (control group) for 16 weeks (32 sessions). The primary outcome is change in Lower Extremity Motor Score (LEMS) from baseline to 16 weeks. Secondary outcomes include walking speed (10-Meter Walk Test), functional independence (Spinal Cord Independence Measure-III), spasticity (Modified Ashworth Scale), electromyography of the lower limb muscles and neurophysiological measures of corticospinal excitability.Expected outcomesWe hypothesize that combined TMS-tSCS will yield superior improvements in LEMS (≥2 points greater improvement) compared to tSCS alone, with enhanced corticospinal tract plasticity as evidenced by neurophysiological measures.Clinical trial registrationClinicalTrials.gov, identifier: NCT07595497.
Posted on

A dual-branch network with brain region-constrained attention for EEG emotion recognition

IntroductionElectroencephalography (EEG)-based emotion recognition provides an objective avenue for affective computing. However, the complexity of EEG signals across temporal, frequency, and spatial domains makes any single dimension inadequate.MethodsTo overcome these limitations, we propose the Brain Region-Constrained Attention Dual-Branch Network (BRAD-Net). This network adopts a parallel Spatio-Temporal and Spectral-Spatial dual-branch architecture to achieve synergistic multi-domain EEG feature learning. Within the spatio-temporal branch, we introduce a novel Brain Region-Constrained Attention mechanism, which strictly confines self-attention computation to channels belonging to the same brain region. This design not only suppresses irrelevant cross-region interference but also incorporates neuroanatomical priors of brain parcellation, thereby enabling effective and interpretable representation learning.ResultsIn subject-dependent experiments using 10-fold cross-validation on DEAP and DREAMER datasets, BRAD-Net achieves high accuracies of 97.44%, 97.70%, and 97.97% for valence, arousal, and dominance on DEAP, and 99.66%, 99.78%, and 99.80% on DREAMER, respectively. Leave-one-subject-out validation on DREAMER dataset achieves accuracies of 72.80% and 75.66% for arousal and dominance, respectively. Additionally, the BRAD-Net demonstrates strong cross-paradigm adaptability, achieving 98.21% accuracy on a depression classification dataset.ConclusionsThese findings confirm that integrating neuroanatomical priors into a dual-branch multi-dimensional learning framework effectively extracts robust and interpretable neural representations. BRAD-Net not only advances high-performance EEG emotion recognition but also provides a novel, biologically-constrained design paradigm for developing more interpretable brain-computer interface models. By demonstrating that restricting attention to within-brain-region interactions suffices for accurate emotion recognition, our work offers a new theoretical perspective on the application of brain parcellation knowledge in classification models.
Posted on

Comparative meta-analysis of task-related functional brain abnormalities in nonsuicidal self-injury and suicide attempt

BackgroundNonsuicidal self-injury (NSSI) and suicide attempt (SA) are two major self-injurious behaviors causing substantial suffering and socioeconomic burden. However, it remains unclear whether NSSI and SA are characterized by common or distinct brain dysregulations. Here, we aimed to identify shared and separable neurofunctional alterations during task engagement between NSSI and SA.MethodA coordinate-based meta-analysis was employed using Seed-based d Mapping with Permutation of Subject Images (SDM-PSI) by capitalizing on task-based functional magnetic resonance imaging (fMRI) studies comparing the brain activation between NSSI/SA individuals and controls.ResultThe search identified 10 studies for NSSI (n = 200, mean age: 22.89 years) and 16 studies for SA (n = 343, mean age: 28.65 years). After threshold-free cluster enhancement correction, NSSI individuals exhibited increased right amygdala activation relative to both controls and the SA group, as well as heightened left middle frontal gyrus and reduced left paracentral lobule activation compared to the SA group. No significant activation differences were found between SA and controls, though a less conservative threshold revealed increased left postcentral gyrus activation in the SA group. No shared functional abnormalities were identified between NSSI and SA under either corrected or uncorrected thresholds. Importantly, subgroup analyses revealed that the neurofunctional abnormalities in NSSI were primarily driven by adolescent cohorts, whereas no significant clusters emerged for the SA group across age-stratified analyses.ConclusionThese results suggest that neurofunctional abnormalities are evident in adolescent NSSI, particularly in fronto−limbic regions, with no robust findings in adult NSSI or SA subgroups. This highlights a unique developmental trajectory that necessitates age-tailored risk assessment and interventions for self-injurious behaviors.
Posted on

The reliability and diagnostic accuracy of the GAD-7 and GAD-2 for the most prevalent anxiety disorders in Latvian primary care

IntroductionAnxiety disorders are common but often under-recognized in primary care. Validation data for the GAD-7 and GAD-2 in Eastern European primary care settings remain limited. To determine the reliability and diagnostic performance of the GAD-7 and GAD-2 for the most prevalent anxiety disorders in Latvian primary care using the Mini International Neuropsychiatric Interview (MINI) as the reference standard.MethodsConsecutive adult patients from 24 primary care practices completed the GAD-7 and GAD-2 prior to consultation. Within two weeks, blinded psychiatrists conducted MINI interviews by telephone. Internal consistency was assessed using Cronbach’s alpha. Diagnostic accuracy was evaluated using receiver operating characteristic analyses, sensitivity, specificity, and likelihood ratios.ResultsThe final analytical sample comprised 1,467 participants. The most prevalent MINI-defined anxiety disorders were agoraphobia (8.0%), generalized anxiety disorder (6.1%), and social phobia (4.7%). Internal consistency was 0.86 for the GAD-7 and 0.72 for the GAD-2. For generalized anxiety disorder, optimal cut-offs were ≥6 (GAD-7; sensitivity 71.9%, specificity 76.3%, AUC = 0.799) and ≥2 (GAD-2; sensitivity 80.9%, specificity 65.8%, AUC = 0.772). GAD-7 and GAD-2 performance for social phobia, agoraphobia and any anxiety disorders was relatively weak. For any anxiety disorder, the optimal thresholds were identified as ≥4 (GAD-7; sensitivity 70.6%, specificity 61.6%, AUC = 0.718) and ≥2 (GAD-2; sensitivity 64.9%, specificity 68.1%, AUC = 0.691).ConclusionThe GAD-7 and GAD-2 demonstrate satisfactory reliability and the strongest diagnostic accuracy to screen for generalized anxiety disorder in Latvian primary care, whereas performance for other anxiety disorder subtypes was comparatively more limited. Subtypes of anxiety disorders should be considered and specific cut-offs calculated to improve the detection of different anxiety disorders in routine general practice.
Posted on

Research progress on addictive features and reward circuit mechanisms in non-suicidal self-injury and the feasibility of precision neuromodulation

Non-Suicidal Self-Injury (NSSI) presents a significant public health challenge; however, its underlying neurobiological mechanisms remain insufficiently understood, limiting the development of targeted interventions. Emerging evidence suggests that NSSI exhibits core addictive features, such as compulsive urges and tolerance, which may be driven by dysfunctions in the brain’s reward circuitry. This review synthesizes current research on the neural overlaps between NSSI and addiction, specifically focusing on the dysregulation of the ventral striatum and prefrontal cortex. Based on this mechanistic framework, we propose the potential of Stanford Accelerated Intelligent Neuromodulation Therapy (SAINT)—a high-dose, functional connectivity-guided transcranial magnetic stimulation protocol—as a precision treatment for NSSI. By targeting specific reward network deficits, SAINT may offer a novel, rapid-acting therapeutic strategy for patients who do not respond to conventional pharmacological or psychological interventions.
Posted on

Self-esteem and inner strengths: a network study in Thai university students with borderline personality disorder symptoms

IntroductionSelf-esteem is widely regarded as an important construct in the psychological functioning of individuals with borderline personality disorder (BPD) symptoms. During emerging adulthood, fluctuations in self-esteem are often linked to emotional dysregulation and maladaptive adjustment; however, self-esteem has rarely been examined within a broader system of culturally relevant psychological resources. Drawing on the Theravāda Buddhist framework of the Ten Pāramīs, inner strengths may represent protective resources that can be examined alongside self-esteem to clarify how these strengths co-occur among students experiencing BPD symptoms.MethodsThe present study employed a regularized psychological network approach to investigate partial associations between self-esteem and Ten Pāramī–based inner strengths among Thai university students screening positive for BPD symptoms. Participants were 346 Thai university students (25.4% male, 74.6% female; mean age = 21.60 ± 2.24 years) identified using a standardized BPD screening instrument. Inner strengths included Truthfulness, Perseverance, Wisdom, Generosity, adherence to the Five Precepts, Meditation, Tolerance, Equanimity, Determination, and Loving-kindness.ResultThe estimated network showed a predominantly positive pattern of partial associations among strengths. The strongest edge was observed between Generosity and Loving-kindness, and links between self-esteem and Determination (and Equanimity) were among the most consistently estimated associations (based on bootstrap confidence intervals). Centrality indices were examined descriptively; however, case-dropping bootstrap results indicated very limited stability of centrality estimates (CS(cor = 0.7) = 0.13 for strength and expected influence), and centrality rankings were therefore treated as strictly exploratory. Discussion: Although some negative partial associations were estimated (e.g., involving Truthfulness, Perseverance, and Equanimity), their precision was limited and such patterns should be treated as exploratory. Exploratory gender-stratified analyses suggested that the strongest edges were similar in the female subsample, whereas the male subsample yielded a sparse/near-empty regularized network, limiting inference regarding gender differences. Given the cross-sectional design, all associations are interpreted as conditional co-occurrence rather than directional or causal effects.ConclusionOverall, these findings highlight a small set of robust co-occurring inner strengths linked to self-esteem in Thai university students with BPD symptoms and provide a culturally informed basis for hypothesis generation regarding strengths-based skills cultivation and supportive interventions in university settings.
Posted on

Hair cortisol as psychotherapy process parameter – an inpatient pediatric psychosomatic study

IntroductionPediatric-psychosomatic inpatient therapy is an essential part of the German health care system for the treatment of mental disorders in children and adolescents. However, empirical research in this field remains scarce and limited to psychological parameters. This longitudinal naturalistic study aimed to evaluate the efficacy and sustainability of inpatient psychosomatic therapy in children and adolescents by examining both psychological outcomes and biological markers.MethodsA total of 58 patients were assessed at seven time points before, during, and after treatment. Hair cortisol concentration (HCC) was measured as a neuroendocrine parameter of long-term stress regulation. Psychometric data were collected using five validated questionnaires.ResultsFindings indicated significant improvements in perceived stress, depressive and anxiety symptoms, family functioning and internalizing symptoms in the course of inpatient treatment. Overall, these effects remained stable at three- and six-month follow-ups, with only transient increases in depressive symptoms and family problems. HCC showed a significant decrease from admission to discharge and remained stable across follow-ups.DiscussionThese results support the efficacy of inpatient pediatric psychosomatic interventions on both psychological outcomes and neuroendocrine stress regulation and highlight the value of integrating biological markers into psychotherapy research.