Generative AI – including large language models such as GPT-4, and image generators such as DALL-E, Midjourney, and Stable Diffusion – is advancing in a "storm of hype and fright", as some commentators have observed. Recent advances in artificial intelligence have yielded warnings that the rapidly developing technology may result in "ever more powerful digital minds that no one – not even their creators – can understand, predict, or reliably control". That's according to an open letter signed by more than 1,000 AI experts, researchers and backers, which calls for an immediate pause on the creation of "giant" AIs for six months so that safety protocols can be developed to mitigate their dangers. But what is the technology currently capable of doing? Midjourney creates images from text descriptions.

It is well understood that modern deep networks are vulnerable to adversarial attacks. However, conventional methods fail to produce adversarial perturbations that are intelligible to humans, and they pose limited threats in the physical world. To study feature-class associations in networks and better understand the realworld threats they face, we develop feature-level adversarial perturbations using deep image generators and a novel optimization objective. We show that they are versatile and use them to generate targeted feature-level attacks at the ImageNet scale that are simultaneously interpretable, universal to any source image, and physically-realizable. These attacks can also reveal spurious, semantically-describable feature/class associations, and we use them to guide the design of "copy/paste" adversaries in which one natural image is pasted into another to cause a targeted misclassification. State-of-the-art neural networks are vulnerable to adversarial inputs, which cause the network to fail yet only differ from benign inputs in subtle ways. Adversaries for visual classifiers conventionally take the form of a small-norm perturbation to a benign source image that causes misclassification (Szegedy et al., 2013; Goodfellow et al., 2014). These are effective, but to a human, these perturbations typically appear as random or mildly-textured noise. As such, analyzing these adversaries does not reveal information about the network relevant to how it will function-and how it may fail-when presented with human-interpretable features. Another limitation with conventional adversaries is that they do not tend to be physically-realizable. While they can retain some effectiveness when printed and photographed in a controlled setting (Kurakin et al., 2016), they are generally ineffective in less controlled ones such as those experienced by autonomous vehicles (Kong et al., 2020). Several works discussed in Section 2 have aimed to produce adversarial modifications that are universal to any source image, interpretable, or physically-realizable.

Future spaceships that humans use to explore other solar systems might be able to fold up and fit into a tiny space, since NASA engineers are exploring foldable designs that could be used on spacecraft today. Smaller and lighter equipment is easier and less costly to launch, making it a more viable option. This strategy has a root in origami, the paper-folding art that has its roots in Japan. "Origami can seem deceptively simple, hiding complex math within its creases," NASA's Jet Propulsion Laboratory said in a statement. "Besides aesthetic beauty, it addresses a persistent problem faced by JPL engineers: How do you pack the greatest amount of spacecraft into the smallest volume possible?"

NASA recently launched Mars rovers that students had designed, sending them into space for about 20 minutes to see how they would perform high above Earth's surface. The students working on the robotics project were from Virginia Tech and the University of Central Florida, NASA reported, and they used 3D printing to build the prototypes that could be used to explore Mars but would store easily, with features like folding or collapsible parts. The rocket last week sent them up 154 miles, then brought them down with a parachute into the Atlantic Ocean. "Part of the problem we keep running into is packaging," NASA research engineer Jamshid Samareh said in the space agency's statement. "We have to carry a lot of payloads -- rovers, habitats and such. We want to package them on top of the launch vehicle. Rovers have classically been bulky and heavy. The students helped design 18 of the more lightweight and less cumbersome Mars rovers, four of which were built and launched on the NASA rocket. "I have always thought of mass to be the limiting factor in space travel," Virginia Tech student Alex Matta, said in the NASA statement. "Participation in this project led me to realize that minimizing volume of the cargo is important as well." The students offered a fresh set of eyes and minds to the effort. Students watch a NASA rocket carry their Mars rover designs into space. NASA research engineer Jamshid Samareh shows off one of the student-made Mars rover designs that recently launched into space. "They come up with these ideas that I cannot come up with," Samareh said. "They have a different mentality.

It costs a stupendous amount of money to send something from the surface of Earth to the surface of Mars, and there are severe limits on the volume and mass that you can send at any one time. In order to stuff the maximum amount of science into the minimum amount of space, NASA has had to get creative, with landers and rovers designed to be lightweight and foldable. At NASA's Jet Propulsion Laboratory, in Pasadena, Calif., engineers have long been trying to cram as much robot as possible into the absolute minimum amount of space, and a team of roboticists there recently showed us their latest creation: PUFFER, the Pop-Up Flat Folding Explorer Robot. It's designed to pack down nearly flat for transport, and then re-expand on site to investigate all the places a bigger rover can't quite reach. The overall idea with PUFFER is that you'd pack a bunch of them along with the next Mars rover, and send them out whenever you want to go somewhere that it would be either risky or impossible for the larger rover to go.

It costs a stupendous amount of money to send something from the surface of Earth to the surface of Mars, and there are severe limits on the volume and mass that you can send at any one time. In order to stuff the maximum amount of science into the minimum amount of space, NASA has had to get creative, with landers and rovers designed to be lightweight and foldable. At NASA's Jet Propulsion Laboratory, in Pasadena, Calif., engineers have long been trying to cram as much robot as possible into the absolute minimum amount of space, and a team of roboticists there recently showed us their latest creation: PUFFER, the Pop-Up Flat Folding Explorer Robot. It's designed to pack down nearly flat for transport, and then re-expand on site to investigate all the places a bigger rover can't quite reach. The overall idea with PUFFER is that you'd pack a bunch of them along with the next Mars rover, and send them out whenever you want to go somewhere that it would be either risky or impossible for the larger rover to go.

NASA may have equipped its Mars Curiosity rover with an impressive array of scientific instruments, but the robot attaché's size and $2.5-billion price tag give its operators ample reason to steer clear of terrain that could jeopardize its mission. Which is a shame, because much of Mars' craggy, cave-ridden, boulder-strewn landscape is so treacherous (planetary geologists literally call it chaos terrain), that big, expensive robots like Curiosity can't risk accessing it. That's why NASA's Jet Propulsion Laboratory built Puffer. Short for Pop-Up Flat Folding Explorer Robot, Puffer is the agency's latest origami-inspired device. JPL has experimented previously with collapsible solar panels based on the Japanese art of folding paper.

March 23, 2017 --If some NASA researchers have their way, Mars exploration technology of the future may rely on an art form from the past. NASA's Jet Propulsion Laboratory (JPL) has developed a Pop-Up Flat Folding Explorer Robot (PUFFER) prototype that could change how we explore Mars. The rugged yet portable machine takes its inspiration from the art of origami, which, despite Americans' association with grade-school arts and crafts, is proving to be a cutting-edge design philosophy. Recent developments in the field have led to an explosion of uses ranging from solar panels to bulletproof barriers. What sets PUFFER apart from other rovers is that it folds flat, making its mini-profile even slimmer.

March 23, 2017 --If some NASA researchers have their way, Mars exploration technology of the future may rely on an art form from the past. NASA's Jet Propulsion Laboratory (JPL) has developed a Pop-Up Flat Folding Explorer Robot (PUFFER) prototype that could change how we explore Mars. The rugged yet portable machine takes its inspiration from the art of origami, which, despite Americans' association with grade-school arts and crafts, is proving to be a cutting-edge design philosophy. Recent developments in the field have led to an explosion of uses ranging from solar panels to bulletproof barriers. What sets PUFFER apart from other rovers is that it folds flat, making its mini-profile even slimmer.

The great explorers Lewis and Clark knew the importance of team expeditions. Now, engineers at NASA's Jet Propulsion Laboratory in Pasadena, California are developing a small scout robot called the Pop-Up Flat Folding Explorer Robot (PUFFER) to accompany the next generation of Martian rovers in their outer space explorations. Inspired by origami, the Japanese art of paper folding, PUFFER is designed to change shape in order to squeeze into small crevasses that are too tight for rovers to reach. So far the two-wheeled scout has been successfully tested in hostile and diverse terrains including the Mojave Desert and Antartica. Though rovers themselves are built to last, they're expensive and NASA engineers take care not to send them on overtly dangerous missions.