OK, now getting officially worried...
"A whistleblower leaks the misalignment memo to the New York Times. For the first time, the public hears about Agent-4. “Secret OpenBrain AI is Out of Control, Insider Warns,” says the headline, and the story goes on to cite evaluations showing off-the-charts bioweapons capabilities, persuasion abilities, the ability to automate most white-collar jobs, and of course the various concerning red flags.
A small study in Israel developed a machine learning model that achieved moderate to high accuracy in predicting romantic attraction and rejection in a simulated dating app scenario. The researchers also found that the model was better able to predict the romantic emotions of picky participants than for those who were less fussy about their romantic interests, suggesting that picky individuals might exhibit stronger, more distinct brain activity related to romantic preferences. The paper was published in Computers in Biology and Medicine. Romantic love is an important aspect of human relationships. For many, it is the most important aspect of their lives. Romantic love promotes intimacy, bonding, and long-term pair connections that support both individual well-being and social stability. Across cultures, romantic love is linked to companionship, emotional security, and the motivation to form families. Psychologically, it engages brain systems tied to reward and attachment, which can foster resilience and life satisfaction. In modern societies, however, the ways people meet potential partners have changed significantly. Dating apps provide a new method for initiating relationships, expanding social networks beyond traditional circles of friends, workplaces, or communities. They can increase efficiency by matching people based on preferences, values, or location, and they offer opportunities to individuals who might otherwise struggle with limited social exposure. At the same time, dating apps can sometimes promote superficial judgments, encourage serial dating, or create unrealistic expectations. Study authors Dor Zazon and Nir Nissim wanted to explore whether machine learning can be used to accurately infer an individual’s romantic emotions based on the analysis of their neural responses to fictional dating profiles using electroencephalography (EEG). They aimed to predict attraction by analyzing event-related potentials (ERPs)—the brain’s specific electrical responses to stimuli like seeing a face—which were captured by the EEG. The 61 participants were students at Ben-Gurion University of the Negev in Israel, aged 23 to 32, and included 31 women. Study authors used 400 free images of people (200 men, 200 women) between 20 and 30 years of age downloaded from the internet to serve as visual stimuli. Each participant was shown 200 of these images (of the gender they declared being attracted to) while their brain activity was recorded using EEG. After viewing an image, the participant had to indicate whether they found the person in the picture attractive or not. In a subsequent experiment, the researchers told participants that their profile picture had been sent to other participants for feedback. They were then shown images of the people they had found attractive, accompanied by feedback indicating whether that person was supposedly attracted to them in return. This feedback, however, was fabricated. The deception was revealed to participants after the experiment concluded. After viewing this feedback, participants had to indicate whether they were happy with it or not (a question designed to ensure engagement). During this time, EEG recordings of their brain activity continued. In this way, the researchers simulated the core feedback loop of attraction and potential rejection common to most dating apps. The researchers tested multiple machine learning models to predict participants’ romantic emotions. The best-performing models achieved accuracy rates significantly above chance. Prediction accuracy was higher for romantic rejection (predicting the brain’s response to the feedback with 81.3% accuracy) than for initial attraction (71.3% accuracy). Results also showed that the machine learning models were more accurate when predicting the responses of picky participants (those who found fewer people attractive) than for those who were less fussy, suggesting that picky individuals may have stronger or more distinct neural signals related to romantic preference. “By analyzing EEG signals, we can predict user actions on dating apps, such as whether they will swipe right or left on a potential match. This prediction can offer insights into user emotions, including whether they find someone physically appealing or experience negative emotions related to rejection,” the study authors concluded. The study contributes to the development of technological ways to decode and interpret human brain responses. However, it should be noted that the study was carried out on a small group of students using static pictures of people. Real dating app profiles usually contain more information than just a picture. Additionally, people use dating apps when they actively try to connect with potential partners, something that might not have been the case with all students participating in this study. Studies of brain responses of individuals fully engaged in searching for a romantic partner might not yield identical results. The paper, “Can your brain signals reveal your romantic emotions?,” was authored by Dor Zazon and Nir Nissim.
A small study in Israel developed a machine learning model that achieved moderate to high accuracy in predicting romantic attraction and rejection in a simulated dating app scenario. The researchers also found that the model was better able to predict the romantic emotions of picky participants than for those who were less fussy about their romantic interests, suggesting that picky individuals might exhibit stronger, more distinct brain activity related to romantic preferences. The paper was published in Computers in Biology and Medicine. Romantic love is an important aspect of human relationships. For many, it is the most important aspect of their lives. Romantic love promotes intimacy, bonding, and long-term pair connections that support both individual well-being and social stability. Across cultures, romantic love is linked to companionship, emotional security, and the motivation to form families. Psychologically, it engages brain systems tied to reward and attachment, which can foster resilience and life satisfaction. In modern societies, however, the ways people meet potential partners have changed significantly. Dating apps provide a new method for initiating relationships, expanding social networks beyond traditional circles of friends, workplaces, or communities. They can increase efficiency by matching people based on preferences, values, or location, and they offer opportunities to individuals who might otherwise struggle with limited social exposure. At the same time, dating apps can sometimes promote superficial judgments, encourage serial dating, or create unrealistic expectations. Study authors Dor Zazon and Nir Nissim wanted to explore whether machine learning can be used to accurately infer an individual’s romantic emotions based on the analysis of their neural responses to fictional dating profiles using electroencephalography (EEG). They aimed to predict attraction by analyzing event-related potentials (ERPs)—the brain’s specific electrical responses to stimuli like seeing a face—which were captured by the EEG. The 61 participants were students at Ben-Gurion University of the Negev in Israel, aged 23 to 32, and included 31 women. Study authors used 400 free images of people (200 men, 200 women) between 20 and 30 years of age downloaded from the internet to serve as visual stimuli. Each participant was shown 200 of these images (of the gender they declared being attracted to) while their brain activity was recorded using EEG. After viewing an image, the participant had to indicate whether they found the person in the picture attractive or not. In a subsequent experiment, the researchers told participants that their profile picture had been sent to other participants for feedback. They were then shown images of the people they had found attractive, accompanied by feedback indicating whether that person was supposedly attracted to them in return. This feedback, however, was fabricated. The deception was revealed to participants after the experiment concluded. After viewing this feedback, participants had to indicate whether they were happy with it or not (a question designed to ensure engagement). During this time, EEG recordings of their brain activity continued. In this way, the researchers simulated the core feedback loop of attraction and potential rejection common to most dating apps. The researchers tested multiple machine learning models to predict participants’ romantic emotions. The best-performing models achieved accuracy rates significantly above chance. Prediction accuracy was higher for romantic rejection (predicting the brain’s response to the feedback with 81.3% accuracy) than for initial attraction (71.3% accuracy). Results also showed that the machine learning models were more accurate when predicting the responses of picky participants (those who found fewer people attractive) than for those who were less fussy, suggesting that picky individuals may have stronger or more distinct neural signals related to romantic preference. “By analyzing EEG signals, we can predict user actions on dating apps, such as whether they will swipe right or left on a potential match. This prediction can offer insights into user emotions, including whether they find someone physically appealing or experience negative emotions related to rejection,” the study authors concluded. The study contributes to the development of technological ways to decode and interpret human brain responses. However, it should be noted that the study was carried out on a small group of students using static pictures of people. Real dating app profiles usually contain more information than just a picture. Additionally, people use dating apps when they actively try to connect with potential partners, something that might not have been the case with all students participating in this study. Studies of brain responses of individuals fully engaged in searching for a romantic partner might not yield identical results. The paper, “Can your brain signals reveal your romantic emotions?,” was authored by Dor Zazon and Nir Nissim.
A small study in Israel developed a machine learning model that achieved moderate to high accuracy in predicting romantic attraction and rejection in a simulated dating app scenario. The researchers also found that the model was better able to predict the romantic emotions of picky participants than for those who were less fussy about their romantic interests, suggesting that picky individuals might exhibit stronger, more distinct brain activity related to romantic preferences. The paper was published in Computers in Biology and Medicine. Romantic love is an important aspect of human relationships. For many, it is the most important aspect of their lives. Romantic love promotes intimacy, bonding, and long-term pair connections that support both individual well-being and social stability. Across cultures, romantic love is linked to companionship, emotional security, and the motivation to form families. Psychologically, it engages brain systems tied to reward and attachment, which can foster resilience and life satisfaction. In modern societies, however, the ways people meet potential partners have changed significantly. Dating apps provide a new method for initiating relationships, expanding social networks beyond traditional circles of friends, workplaces, or communities. They can increase efficiency by matching people based on preferences, values, or location, and they offer opportunities to individuals who might otherwise struggle with limited social exposure. At the same time, dating apps can sometimes promote superficial judgments, encourage serial dating, or create unrealistic expectations. Study authors Dor Zazon and Nir Nissim wanted to explore whether machine learning can be used to accurately infer an individual’s romantic emotions based on the analysis of their neural responses to fictional dating profiles using electroencephalography (EEG). They aimed to predict attraction by analyzing event-related potentials (ERPs)—the brain’s specific electrical responses to stimuli like seeing a face—which were captured by the EEG. The 61 participants were students at Ben-Gurion University of the Negev in Israel, aged 23 to 32, and included 31 women. Study authors used 400 free images of people (200 men, 200 women) between 20 and 30 years of age downloaded from the internet to serve as visual stimuli. Each participant was shown 200 of these images (of the gender they declared being attracted to) while their brain activity was recorded using EEG. After viewing an image, the participant had to indicate whether they found the person in the picture attractive or not. In a subsequent experiment, the researchers told participants that their profile picture had been sent to other participants for feedback. They were then shown images of the people they had found attractive, accompanied by feedback indicating whether that person was supposedly attracted to them in return. This feedback, however, was fabricated. The deception was revealed to participants after the experiment concluded. After viewing this feedback, participants had to indicate whether they were happy with it or not (a question designed to ensure engagement). During this time, EEG recordings of their brain activity continued. In this way, the researchers simulated the core feedback loop of attraction and potential rejection common to most dating apps. The researchers tested multiple machine learning models to predict participants’ romantic emotions. The best-performing models achieved accuracy rates significantly above chance. Prediction accuracy was higher for romantic rejection (predicting the brain’s response to the feedback with 81.3% accuracy) than for initial attraction (71.3% accuracy). Results also showed that the machine learning models were more accurate when predicting the responses of picky participants (those who found fewer people attractive) than for those who were less fussy, suggesting that picky individuals may have stronger or more distinct neural signals related to romantic preference. “By analyzing EEG signals, we can predict user actions on dating apps, such as whether they will swipe right or left on a potential match. This prediction can offer insights into user emotions, including whether they find someone physically appealing or experience negative emotions related to rejection,” the study authors concluded. The study contributes to the development of technological ways to decode and interpret human brain responses. However, it should be noted that the study was carried out on a small group of students using static pictures of people. Real dating app profiles usually contain more information than just a picture. Additionally, people use dating apps when they actively try to connect with potential partners, something that might not have been the case with all students participating in this study. Studies of brain responses of individuals fully engaged in searching for a romantic partner might not yield identical results. The paper, “Can your brain signals reveal your romantic emotions?,” was authored by Dor Zazon and Nir Nissim.
A small study in Israel developed a machine learning model that achieved moderate to high accuracy in predicting romantic attraction and rejection in a simulated dating app scenario. The researchers also found that the model was better able to predict the romantic emotions of picky participants than for those who were less fussy about their romantic interests, suggesting that picky individuals might exhibit stronger, more distinct brain activity related to romantic preferences. The paper was published in Computers in Biology and Medicine. Romantic love is an important aspect of human relationships. For many, it is the most important aspect of their lives. Romantic love promotes intimacy, bonding, and long-term pair connections that support both individual well-being and social stability. Across cultures, romantic love is linked to companionship, emotional security, and the motivation to form families. Psychologically, it engages brain systems tied to reward and attachment, which can foster resilience and life satisfaction. In modern societies, however, the ways people meet potential partners have changed significantly. Dating apps provide a new method for initiating relationships, expanding social networks beyond traditional circles of friends, workplaces, or communities. They can increase efficiency by matching people based on preferences, values, or location, and they offer opportunities to individuals who might otherwise struggle with limited social exposure. At the same time, dating apps can sometimes promote superficial judgments, encourage serial dating, or create unrealistic expectations. Study authors Dor Zazon and Nir Nissim wanted to explore whether machine learning can be used to accurately infer an individual’s romantic emotions based on the analysis of their neural responses to fictional dating profiles using electroencephalography (EEG). They aimed to predict attraction by analyzing event-related potentials (ERPs)—the brain’s specific electrical responses to stimuli like seeing a face—which were captured by the EEG. The 61 participants were students at Ben-Gurion University of the Negev in Israel, aged 23 to 32, and included 31 women. Study authors used 400 free images of people (200 men, 200 women) between 20 and 30 years of age downloaded from the internet to serve as visual stimuli. Each participant was shown 200 of these images (of the gender they declared being attracted to) while their brain activity was recorded using EEG. After viewing an image, the participant had to indicate whether they found the person in the picture attractive or not. In a subsequent experiment, the researchers told participants that their profile picture had been sent to other participants for feedback. They were then shown images of the people they had found attractive, accompanied by feedback indicating whether that person was supposedly attracted to them in return. This feedback, however, was fabricated. The deception was revealed to participants after the experiment concluded. After viewing this feedback, participants had to indicate whether they were happy with it or not (a question designed to ensure engagement). During this time, EEG recordings of their brain activity continued. In this way, the researchers simulated the core feedback loop of attraction and potential rejection common to most dating apps. The researchers tested multiple machine learning models to predict participants’ romantic emotions. The best-performing models achieved accuracy rates significantly above chance. Prediction accuracy was higher for romantic rejection (predicting the brain’s response to the feedback with 81.3% accuracy) than for initial attraction (71.3% accuracy). Results also showed that the machine learning models were more accurate when predicting the responses of picky participants (those who found fewer people attractive) than for those who were less fussy, suggesting that picky individuals may have stronger or more distinct neural signals related to romantic preference. “By analyzing EEG signals, we can predict user actions on dating apps, such as whether they will swipe right or left on a potential match. This prediction can offer insights into user emotions, including whether they find someone physically appealing or experience negative emotions related to rejection,” the study authors concluded. The study contributes to the development of technological ways to decode and interpret human brain responses. However, it should be noted that the study was carried out on a small group of students using static pictures of people. Real dating app profiles usually contain more information than just a picture. Additionally, people use dating apps when they actively try to connect with potential partners, something that might not have been the case with all students participating in this study. Studies of brain responses of individuals fully engaged in searching for a romantic partner might not yield identical results. The paper, “Can your brain signals reveal your romantic emotions?,” was authored by Dor Zazon and Nir Nissim.
HyperAgents: Self-referential self-improving agents
https://github.com/facebookresearch/hyperagents
#HackerNews #HyperAgents #HyperAgentsResearch #SelfImprovingAgents #AIResearch #FacebookResearch
Andrej Karpathy's recent podcast interview is worth your time
Key ideas: agent orchestration over single-session prompting,
AutoResearch loops that remove human researcher from hyperparameter tuning, and a prediction that digital transformation leads while physical robotics lags by years
His take on open source (~6-8 months behind frontier) being a healthy power balance is worth sitting with
"Centralization has a very poor track record." Hard to argue
https://winbuzzer.com/2026/03/26/tencent-open-sources-covo-audio-7b-speech-language-model-xcxwbn/
Tencent Releases Covo-Audio Open-Source 7B Speech AI Model
#AI #Tencent #TencentAILab #CovoAudio #AIModels #AIResearch #AIAudio #AIVoiceGeneration #OpenSourceAI #SpeechSynthesis
𝙳𝚊𝚒𝚕𝚢 𝙵𝚒𝚜𝚑𝚠𝚛𝚊𝚙
AIagent.at 🤖 AI News
purescience_news
purescience_news
Hacker News
sidj79
Winbuzzer