'The mouse is long overdue for reinvention.' Breakthroughs, discoveries, and DIY tips sent every weekday. Many of us subscribe to the old adage, "If it ain't broke, don't fix it." But what if that something was actually broken all along and we just didn't realize it? That's the argument presented in an upcoming issue of the journal by researchers from Nazarbayev University in Kazakhstan.

The first human patient implanted with a brain-chip from Neuralink appears to have fully recovered and is able to control a computer mouse using their thoughts, the startup's founder, Elon Musk, said late on Monday. "Progress is good, and the patient seems to have made a full recovery, with no ill effects that we are aware of. Patient is able to move a mouse around the screen by just thinking," Musk said in a Spaces event on the social media platform X. Musk said Neuralink was now trying to get as many mouse button clicks as possible from the patient. Neuralink did not immediately reply to a request for further details. The firm successfully implanted a chip on its first human patient last month, after receiving approval for human trial recruitment in September.

Once upon a time, long before smartphones or even laptops were ubiquitous, the computer mouse was new, and it was thrilling. The 1984 Macintosh wasn't the first machine to come with one, but it was the first to popularize the gizmo for ordinary people. Proper use of the mouse was not intuitive. Many people had a hard time moving and clicking at the same time, and "double-clicking" was a skill one had to learn. Still, anyone could put a hand on the thing, move it around on a table, and see the results on-screen: A little cursor moved along with you.

A computer mouse that inspired Steve Jobs to create Apple's first mice has clicked with bidders at a $178,936 (£147,000) auction. The early mouse and coding keyset was created by computing legend Douglas Engelbart, a pioneer of the controller system. The lot sold for approximately 12 times its estimate of $14,640 (£12,000) in a sale by Boston-based RR Auction on Thursday. The rare, early three-button computer mouse designed by Engelbart utilises two metal discs which correspond to the X-axis and Y-axis on the bottom to locate the position of the cursor, rather than a ball or optical light that came to be used later. The coding keyset, features five keys, permitting 31 key-press combinations, for typing and entering commands.

Robotics and artificial intelligence are beginning to transform a broad spectrum of industries, and I believe that in time, given its near-endless applications, omnipresent AI will contribute $150 trillion to the global economy. I also believe that humanitarian efforts all over the world will be greatly advanced by emerging artificial intelligence and robotics technologies as they continue on their rapid exponential growth trajectory. But before we can begin to analyse and realize the potential of humanitarian AI and robotics, we have to address the most common questions that people ask about the future of human and machine collaboration. Will robots take our jobs? And will artificial intelligence and robots create a more unequal society?

Imagine visiting the control room of a metals company. You're there to discuss asset performance and process optimization. During the visit, you see on a desk a computer mouse wrapped in a rubber band. On the nearby computer screen, the cursor hovers over an icon that a person would click to acknowledge an alarm triggered by the automated system tracking the thousands of sensors placed throughout the company's facilities. It seems it would never be clear to the person sitting in that chair if there was a serious problem or not.

Recent work has thoroughly documented the susceptibility of deep learning systems to adversarial examples, but most such instances directly manipulate the digital input to a classifier. Although a smaller line of work considers physical adversarial attacks, in all cases these involve manipulating the object of interest, e.g., putting a physical sticker on a object to misclassify it, or manufacturing an object specifically intended to be misclassified. In this work, we consider an alternative question: is it possible to fool deep classifiers, over all perceived objects of a certain type, by physically manipulating the camera itself? We show that this is indeed possible, that by placing a carefully crafted and mainly-translucent sticker over the lens of a camera, one can create universal perturbations of the observed images that are inconspicuous, yet reliably misclassify target objects as a different (targeted) class. To accomplish this, we propose an iterative procedure for both updating the attack perturbation (to make it adversarial for a given classifier), and the threat model itself (to ensure it is physically realizable). For example, we show that we can achieve physically-realizable attacks that fool ImageNet classifiers in a targeted fashion 49.6% of the time. This presents a new class of physically-realizable threat models to consider in the context of adversarially robust machine learning. Our demo video can be viewed at: https://youtu.be/wUVmL33Fx54

Dr. Engelbart died on Tuesday at 88 at his home in Atherton, Calif. His wife, Karen O'Leary Engelbart, said the cause was kidney failure. Computing was in its infancy when Dr. Engelbart entered the field. Computers were ungainly room-size calculating machines that could be used by only one person at a time. Someone would feed them information in stacks of punched cards and then wait hours for a printout of answers. Interactive computing was a thing of the future, or in science fiction.

Polysemy is a problem for methods that exploit image search engines to build object category models. Existing unsupervised approaches do not take word sense into consideration. We propose a new method that uses a dictionary to learn models of visual word sense from a large collection of unlabeled web data. The use of LDA to discover a latent sense space makes the model robust despite the very limited nature of dictionary definitions. The definitions are used to learn a distribution in the latent space that best represents a sense. The algorithm then uses the text surrounding image links to retrieve images with high probability of a particular dictionary sense. An object classifier is trained on the resulting sense-specific images. We evaluate our method on a dataset obtained by searching the web for polysemous words. Category classification experiments show that our dictionary-based approach outperforms baseline methods.