Growing evidence suggests that the brain uses an "attention schema" to monitor, predict, and help control attention. It has also been suggested that an attention schema improves social intelligence by allowing one person to better predict another. Given their potential advantages, attention schemas have been increasingly tested in machine learning. Here we test small deep learning networks to determine how the addition of an attention schema may affect performance on a range of tasks. First, we found that an agent with an attention schema is better at judging or categorizing the attention states of other agents. Second, we found that an agent with an attention schema develops a pattern of attention that is easier for other agents to judge and categorize. Third, we found that in a joint task where two agents paint a scene together and must predict each other's behavior for best performance, adding an attention schema improves that performance. Finally, we find that the performance improvements caused by an attention schema are not a non-specific result of an increase in network complexity. Not all performance, on all tasks, is improved. Instead, improvement is specific to "social" tasks involving judging, categorizing, or predicting the attention of other agents. These results suggest that an attention schema may be useful in machine learning for improving cooperativity and social behavior.

Self-models have been a topic of great interest for decades in studies of human cognition and more recently in machine learning. Yet what benefits do self-models confer? Here we show that when artificial networks learn to predict their internal states as an auxiliary task, they change in a fundamental way. To better perform the self-model task, the network learns to make itself simpler, more regularized, more parameter-efficient, and therefore more amenable to being predictively modeled. To test the hypothesis of self-regularizing through self-modeling, we used a range of network architectures performing three classification tasks across two modalities. In all cases, adding self-modeling caused a significant reduction in network complexity. The reduction was observed in two ways. First, the distribution of weights was narrower when self-modeling was present. Second, a measure of network complexity, the real log canonical threshold (RLCT), was smaller when self-modeling was present. Not only were measures of complexity reduced, but the reduction became more pronounced as greater training weight was placed on the auxiliary task of self-modeling. These results strongly support the hypothesis that self-modeling is more than simply a network learning to predict itself. The learning has a restructuring effect, reducing complexity and increasing parameter efficiency. This self-regularization may help explain some of the benefits of self-models reported in recent machine learning literature, as well as the adaptive value of self-models to biological systems. In particular, these findings may shed light on the possible interaction between the ability to model oneself and the ability to be more easily modeled by others in a social or cooperative context.

From the classic fingers pinched against thumbs to indicate frustration, bewilderment or delight to airy waves and shrugs, the Italian obsession with gestures is legendary. Now, a study has confirmed that Italians really do talk with their hands. Passionate people from Italy use more than 40 hand gestures a minute while speaking, the study revealed. That's twice what an average Swede does, according to researchers from Lund University. And while other cultures use gestures to help illustrate parts of a story, Italians use them as a kind of running commentary on what they're saying.

Attention has become a common ingredient in deep learning architectures. It adds a dynamical selection of information on top of the static selection of information supported by weights. In the same way, we can imagine a higher-order informational filter built on top of attention: an Attention Schema (AS), namely, a descriptive and predictive model of attention. In cognitive neuroscience, Attention Schema Theory (AST) supports this idea of distinguishing attention from AS. A strong prediction of this theory is that an agent can use its own AS to also infer the states of other agents' attention and consequently enhance coordination with other agents. As such, multi-agent reinforcement learning would be an ideal setting to experimentally test the validity of AST. We explore different ways in which attention and AS interact with each other. Our preliminary results indicate that agents that implement the AS as a recurrent internal control achieve the best performance. In general, these exploratory experiments suggest that equipping artificial agents with a model of attention can enhance their social intelligence.

A Princeton neuroscientist has warned that artificial intelligence-powered chatbots such as ChatGPT are sociopaths without the one thing that makes humans special. In a new essay detailed by The Wall Street Journal, Princeton neuroscientist Michael Graziano explains that AI-powered chatbots are sociopaths without consciousness and that until developers can implement consciousness, they will pose a real danger to humans. For those that don't know, AI chatbots such as ChatGPT are designed to have human-like conversations by remembering what was written by the human earlier in the conversation, providing almost real-time answers and thorough answers to questions. While the dangers of AI aren't so prevalent now, in the future, that could very well change as these sophisticated tools are further upgraded and developed. In order to make them more human-like, Graziano proposes that they are taught human traits such as empathy and prosocial behavior. Notably, the neuroscientist says that these systems will need a form of implemented consciousness to understand these traits and, in turn, adjust their responses to align more with human values.

Without consciousness, Princeton neuroscientist Michael Graziano warns in a new essay published by The Wall Street Journal, artificial intelligence-powered chatbots are doomed to be dangerous sociopaths that could pose a real danger to human beings. With the rise of chatbots like ChatGPT, powerful systems that can imitate the human mind to an impressive degree, AI tools have become more accessible than ever before. But those algorithms will glibly fib about anything that suits their purpose. To make align them with our values, Graziano thinks, they're going to need consciousness. "Consciousness is part of the tool kit that evolution gave us to make us an empathetic, prosocial species," Graziano writes.

When it comes to artificial intelligence, the most important thing is to start investing now, says W.K. Kellogg Foundation's investments director, Neal Graziano. And that's exactly what his organization is doing, along with the Robert Wood Johnson Foundation, institutional investor network Trusted Insight, and European family offices. Together, the groups have directed a $27 million investment, announced last week, into a German venture capital firm that seeds machine learning companies. Founded in 2016, the Berlin-based Merantix is a lean operation. But it has already built three European companies, including Vara, which is Germany's first approved AI software for cancer screening and is touted to have a better accuracy rate than radiology.

Craig Federighi, Apple's senior vice president of software engineering says there are two things you can do to stop nefarious actors from forcing you into FaceID. According to Federighi, "If you don't stare at the phone, it won't unlock," & "If you grip the buttons on both sides of the phone when you hand it over, it will temporarily disable Face ID." Clearly, iPhone X owners will have to practice their squeezing techniques. It would be painful and costly to be held up and discover that you were squeezing it all wrong. The ACLU & the EFF recently sued the DHS for searching the phones and laptops of 11 plaintiffs at the US border without a warrant. The group of plaintiffs includes 10 US citizens and one lawful permanent resident, several of whom are Muslims or people of color.