Watch this video on Streamable.
Streamable Video
Posted in futurism
The last decade has marked a profound change in how we perceive and talk about Artificial Intelligence. The concept of learning, once confined in the corner of AI, has now become so important some people came up with the new term “Machine Intelligence”[1][2][3] as to make clear the fundamental role of Machine Learning in it and further depart form older symbolic approaches.
Recent Deep Learning (DL) techniques have literally swept away previous AI approaches and have shown how beautiful, end-to-end differentiable functions can be learned to solve incredibly complex tasks involving high-level perception abilities.
Yet, since DL techniques have been proven shining only with a large number of labeled examples, the research community has now shifted his attention towards Unsupervised and Reinforcement Learning, both aiming to solve equivalently complex tasks but without (or less as possible) explicit supervision.
Researchers built an AI that gives out money based on who started with less resources—and humans preferred it.
Summary: Bergmann glial cell synaptic engulfing in the cerebellum was enhanced during motor learning in mice.
Source: Tohoku University.
Tohoku University researchers have shown that Bergmann glial cells, astrocyte-like cells in the cerebellum, ‘eat’ their neighboring neuronal elements within healthy living brain tissue.
Using non-invasive techniques to manipulate our emotions, it might be possible to curtail the screaming horrors that plague our sleep.
A study conducted on 36 patients diagnosed with a nightmare disorder showed that a combination of two simple therapies reduced the frequency of their bad dreams.
Scientists invited the volunteers to rewrite their most frequent nightmares in a positive light and then playing sound associated with positive experiences as they slept.
Spaceship Neptune is the latest space tourism venture for a company is planning balloon flights to the edge of space.
Although these studies collectively suggest that the DLPFC plays a major role in making risky choices, a question remains as to whether its activity mediates risky choice via probability weighting, via marginal utility (value) of monetary outcomes, or both.
In the present study we causally address the hypothesis that the DLPFC is involved both in the subjective valuation of a monetary reward and in probability weighting. The hypothesis was not preregistered but was formulated prior to the collection of the data. The hypothesis was well grounded in the existing literature on the role of DLPFC in risk taking. Several previous studies mentioned the possible role of the lateral PFC in separate components of choice under risk, such as reward magnitude, reward probability, and expected value14,25,26. However, previous studies including those exploring causal role of the DLPFC in risky choice with non-invasive brain stimulation were not focusing on the estimation of the risk preference parameters but rather on observing changes purely on a behavioural level. Therefore, in the present study we used an experimental design that is typically employed in economic studies estimating risk preference parameters27. We combined offline repetitive TMS over the left and right DLPFC and sham over the right DLPFC, performed in a randomized and counterbalanced order, with a random lottery pair (RLP) task, which is widely used in economics to estimate the degree of risk aversion as well as the curvature of the probability weighting function on an individual level.
Following offline TMS, subjects completed a computerized task consisting of 96 binary lottery choice questions presented in random order. Using the hierarchical Bayesian modeling approach, we then estimated the structural parameters of risk preferences (degree of risk aversion and the curvature of the probability weighting function) and analyzed the obtained posterior distributions to determine the effect of stimulation on model parameters.
A new tool can determine whether a collection of building blocks will assemble into a mechanically sound structure.
Using the IceCube Neutrino Observatory in Antarctica, researchers have found significant evidence of a cosmic source of high-energy neutrinos.
In 1911, physicist Heike Kamerlingh Onnes used liquid helium—whose production method he invented—to cool mercury to a few kelvins, discovering that its electrical resistance dropped to nil. Although mercury was later found to be a “conventional” superconductor, no microscopic theory so far managed to fully explain the metal’s behavior and to predict its critical temperature TC. Now, 111 years after Kamerlingh Onnes’ discovery, theorists have done just that. Their first-principles calculations accurately predict mercury’s TC but also pinpoint theoretical caveats that could inform searches for room-temperature superconductors [1].
Mercury is an exception among conventional superconductors, most of which can be successfully described with state-of-the-art density-functional-theory methods. To tackle mercury’s unique challenges, Gianni Profeta of the University of L’Aquila, Italy, and colleagues scrutinized all physical properties relevant for conventional superconductivity, which is mediated by the coupling of electrons to phonons. In particular, the researchers accounted for previously neglected relativistic effects that alter phonon frequencies, they improved the description of electron-correlation effects that modify electronic bands, and they showed that mercury’s d-electrons provide an anomalous screening effect that promotes superconductivity by reducing Coulomb repulsion between superconducting electrons. With these improvements, their calculations delivered a TC prediction for mercury only 2.5% lower than the experimental value.
The new understanding of the oldest superconductor will find a place in textbooks but may also offer valuable lessons for superconductivity research, says Profeta. A promising material-by-design approach involves “high-throughput” computations that screen millions of theoretical material combinations to suggest those that could be conventional superconductors close to ambient conditions. “If we don’t include subtle effects similar to those relevant for mercury, these computations may overlook many interesting materials or err in their critical temperature predictions by hundreds of kelvins,” he says.