AI & Robotics News and Discussions

[Yuli Ban]

SoftBank shrinks robotics business, stops Pepper production

SoftBank Group Corp is slashing jobs at its global robotics business and has stopped producing its Pepper robot, according to sources and documents reviewed by Reuters, as the conglomerate downgrades its industry ambitions.

Production of the humanoid Pepper, touted as the first robot with "a heart", was stopped last year, according to three sources familiar with the matter and the documents. It would be costly to restart production, two of the sources said.

Built by Foxconn (2317.TW) in China, Pepper was meant to help plug labour shortages but struggled to find a global customer base. Only 27,000 were produced, one of the sources said.

The pullback reflects the fading of Chief Executive Masayoshi Son's plan to make SoftBank the leader in the robotics industry, producing human-like machines that could serve customers and babysit kids.

Slightly disappointing because I liked Pepper's sleek design. However, it's clear that SoftBank isn't going to lead this field. They already lost Boston Dynamics, and as much as I liked the aesthetics of Pepper... it's really the last straggler of that embarrassing wave of social robots of the 2010s. The entire point of it being "emotional" left it doomed from the start because it's using 2010s-era chatbot technology.

Those social robots would actually stand a chance this coming decade, actually. If Pepper could use LaMDA instead of Watson, it'd have gone so much further. The technology is almost there. All these companies just started it off too soon.

[waitingforthe2030s]

Microsoft and OpenAI have a new A.I. tool that will give coding suggestions to software developers

Microsoft on Tuesday announced an artificial intelligence system that can recommend code for software developers to use as they write code.

Microsoft is looking to simplify the process of programming, the area where the company got its start in 1975. That could keep programmers who already use the company’s tools satisfied and also attract new ones.

The system, called GitHub Copilot, draws on source code uploaded to code-sharing service GitHub, which Microsoft acquired in 2018, as well as other websites. Microsoft and GitHub developed it with help from OpenAI, an AI research start-up that Microsoft backed in 2019.

Researchers at Microsoft and other institutions have been trying to teach computers to write code for decades. The concept has yet to go mainstream, at times because programs to write programs have not been versatile enough. The GitHub Copilot effort is a notable attempt in the field, relying as it does on a large volume of code in many programming languages and vast Azure cloud computing power.

[Yuli Ban]

[Yuli Ban]

[Yuli Ban]

AI Designs Quantum Physics Experiments Beyond What Any Human Has Conceived
Originally built to speed up calculations, a machine-learning system is now making shocking progress at the frontiers of experimental quantum physics

Quantum physicist Mario Krenn remembers sitting in a café in Vienna in early 2016, poring over computer printouts, trying to make sense of what MELVIN had found. MELVIN was a machine-learning algorithm Krenn had built, a kind of artificial intelligence. Its job was to mix and match the building blocks of standard quantum experiments and find solutions to new problems. And it did find many interesting ones. But there was one that made no sense.

“The first thing I thought was, ‘My program has a bug, because the solution cannot exist,’” Krenn says. MELVIN had seemingly solved the problem of creating highly complex entangled states involving multiple photons (entangled states being those that once made Albert Einstein invoke the specter of “spooky action at a distance”). Krenn and his colleagues had not explicitly provided MELVIN the rules needed to generate such complex states, yet it had found a way. Eventually, he realized that the algorithm had rediscovered a type of experimental arrangement that had been devised in the early 1990s. But those experiments had been much simpler. MELVIN had cracked a far more complex puzzle.

[caltrek]

Robots Were Supposed to Take Our Jobs. Instead, They’re Making Them Worse.
by Emily Stewart
July 2, 2021

https://www.vox.com/the-goods/22557895/ ... -uber-lyft

Introduction:

(Vox) The robot revolution is always allegedly just around the corner. In the utopian vision, technology emancipates human labor from repetitive, mundane tasks, freeing us to be more productive and take on more fulfilling work. In the dystopian vision, robots come for everyone’s jobs, put millions and millions of people out of work, and throw the economy into chaos.

Such a warning was at the crux of Andrew Yang’s ill-fated presidential campaign, helping propel his case for universal basic income that he argued would become necessary when automation left so many workers out. It’s the argument many corporate executives make whenever there’s a suggestion they might have to raise wages: $15 an hour will just mean machines taking your order at McDonald’s instead of people, they say. It’s an effective scare tactic for some workers.

But we often spend so much time talking about the potential for robots to take our jobs that we fail to look at how they are already changing them — sometimes for the better, but sometimes not. New technologies can give corporations tools for monitoring, managing, and motivating their workforces, sometimes in ways that are harmful. The technology itself might not be innately nefarious, but it makes it easier for companies to maintain tight control on workers and squeeze and exploit them to maximize profits.

[Yuli Ban]

Nvidia benchmark tests show impressive gains in training AI models
Nvidia announced that systems based on its graphics processor units (GPUs) are delivering 3 to 5 times better performance when it comes to training AI models than they did a year ago, according to the latest MLPerf benchmarks published yesterday

Nvidia announced that systems based on its graphics processor units (GPUs) are delivering 3 to 5 times better performance when it comes to training AI models than they did a year ago, according to the latest MLPerf benchmarks published yesterday.

The MLPerf benchmark is maintained by the MLCommons Association, a consortium backed by Alibaba, Facebook AI, Google, Intel, Nvidia, and others that acts as an independent steward.

The latest set of benchmarks span eight different workloads covering a range of use cases for AI model training, including speech recognition, natural language processing, object detection, and reinforcement learning. Nvidia claims its OEM partners were the only systems vendors to run all the workloads defined by the MLPerf benchmark across a total of 4,096 GPUs. Dell, Fujitsu, Gigabyte Technology, Inspur, Lenovo, Nettrix, and Supermicro all provided on-premises systems certified by Nvidia that were used to run the benchmark.

[Yuli Ban]

OpenAI Launches GitHub Copilot: AI Focused On Code Generation. Should We Be Worried Now?

What can a $1 billion investment buy?
On Tuesday this week, OpenAI and GitHub answered this question boldly with the preview of a new AI tool — GitHub Copilot. It can write user-compatible code and is much better at the task than its predecessor — GPT-3.
Copilot autocompletes code snippets, suggests new lines of code, and can even write whole functions based on the description provided. According to the GitHub blog, the tool is not just a language-generating algorithm based on user input — it is a virtual pair programmer.
It learns and adapts to the user’s coding habits, analyzes the available codebase, and generates suggestions backed by billions of lines of public code it has been trained on.

[Yuli Ban]

How MUM improved Google Searches for vaccine information [Google has already been using MUM to improve search, apparently.]

Soda, pop; sweater, jumper; soccer, football. So many things go by different names. Sometimes it’s a function of language, but sometimes it’s a matter of cultural trends or nuance, or simply where you are in the world.

One very relevant example is COVID-19. As people everywhere searched for information, we had to learn to identify all the different phrases people used to refer to the novel coronavirus to make sure we surfaced high quality and timely information from trusted health authorities like the World Health Organization and Centers for Disease Control and Prevention. A year later, we’re encountering a similar challenge with vaccine names, only this time, we have a new tool to help: Multitask Unified Model (MUM).

[caltrek]

Underwater Robot May Unearth Climate Mysteries
by Claire Hogan
July 2, 2021

https://www.sciencemag.org/news/2021/07 ... -mysteries

Introduction:

(Science) Two hundred meters under the sunny waves of the ocean lies the mesopelagic zone, a cold, dark section of water where humans rarely venture. This area, dubbed the “twilight zone,” houses animals like krill, squid, and jellyfish.

Twilight zone animals play a major part in the carbon cycle, bringing organic carbon from surface waters and trapping them deep beneath the tides. But these shy creatures are delicate and hard to observe, making it nearly impossible to trace their movements, let alone their impact on Earth’s climate.

Enter Mesobot. This autonomous underwater robot weighs in at about 250 kilograms, with a black-and-yellow tanklike exterior. It can track a single organism for more than 1 day without human intervention, relying on a long-lasting battery and advanced tracking algorithms to follow creatures on their daily commutes. With an array of sensors and a high-definition camera, Mesobot could help scientists learn about this mysterious ocean area—and the creatures it contains.

[caltrek]

Merger of Robotic Spine Surgery Businesses
by Brian Heater
July 8, 2021

https://techcrunch.com/2021/07/08/robotics-roundup-19/

Extract:

(TechCrunch)...robotic surgery company Fusion Robotics announced this week announced plans to merge with Adaptive Geometry, another tech company specializing in spinal surgery technology. The two companies will combine to create the perfectly nondescript Accelus (frankly, Fusion is a pretty good name for two combined companies, but maybe that’s just me).

“Accelus will create opportunities for wide-scale adoption of robotics in spine surgery—both in hospitals and ambulatory surgery centers (ASCs)—by addressing previous constraints related to cost and efficiency,” Accelus Chris Walsh said in a release. “Both Fusion Robotics and Integrity Implants have built enabling technology platforms that create a force multiplier for spinal care. Our products and culture create accessibility to fit each patient’s anatomy, each surgeon’s preferred approach, and each healthcare facility’s space and budget limitations, embodying our core principle of access without compromise.”

[Yuli Ban]

Thousands of galaxies classified in the blink of an eye
Astronomers have designed and trained AI that can classify tens of thousands of galaxies in just a few seconds, a task that usually takes months to accomplish

Astronomers have designed and trained a computer program that can classify tens of thousands of galaxies in just a few seconds, a task that usually takes months to accomplish.

In research published today, astrophysicists from Australia have used machine learning to speed up a process that is often done manually by astronomers and citizen scientists around the world.

"Galaxies come in different shapes and sizes," said lead author Mitchell Cavanagh, a Ph.D. candidate based at the University of Western Australia node of the International Centre for Radio Astronomy Research (ICRAR).

"Classifying the shapes of galaxies is an important step in understanding their formation and evolution, and can even shed light on the nature of the Universe itself."

[Yuli Ban]

Paris welcomes first pizzeria operated entirely by robots

A new kind of chef is making a splash on the Parisian dining scene in the Beaubourg neighbourhood of the french capital.

The new pizzeria ‘Pazzi’ is staffed entirely by robots, from order-taking to prepping the dough

to putting the pizzas in boxes.

"We are in a very fast process, with a perfect control of time, a control of quality since we have a constancy offered by robotics, and then an environment that is quite cool and relaxed” says the co-inventor of the ‘Pazzi’ robot, Sébastien Roverso.

“The idea is also to spend a few pleasant minutes watching the robot while you wait for your pizza to be made."

This photograph taken on July 1, 2021, shows "Pazzi", a pizza-making robot at work in a restaurant in Paris. BERTRAND GUAY/AFP

[Yuli Ban]

USC researchers enable AI to use its "imagination"
USC research team has developed an AI that uses human like capabilities to imagine a never before seen object with different attributes, a technique could lead to fairer AI, new medicines and increased autonomous vehicle safety

Imagine an orange cat. Now, imagine the same cat, but with coal-black fur. Now, imagine the cat strutting along the Great Wall of China. Doing this, a quick series of neuron activations in your brain will come up with variations of the picture presented, based on your previous knowledge of the world.

In other words, as humans, it’s easy to envision an object with different attributes. But, despite advances in deep neural networks that match or surpass human performance in certain tasks, computers still struggle with the very human skill of “imagination.”

Now, a USC research team comprising computer science Professor Laurent Itti, and PhD students Yunhao Ge, Sami Abu-El-Haija and Gan Xin, has developed an AI that uses human-like capabilities to imagine a never-before-seen object with different attributes. The paper, titled Zero-Shot Synthesis with Group-Supervised Learning, was published in the 2021 International Conference on Learning Representations on May 7.

“We were inspired by human visual generalization capabilities to try to simulate human imagination in machines,” said Ge, the study’s lead author.

“Humans can separate their learned knowledge by attributes—for instance, shape, pose, position, color—and then recombine them to imagine a new object. Our paper attempts to simulate this process using neural networks.”

IN A NEW APPROACH TO TEACHING AI’S TO “IMAGINE THE UNSEEN,” TRAINING IMAGES (BOTTOM) ARE COMBINED TO SYNTHESIZE THE REQUESTED IMAGE (TOP). IMAGE/GE ET AL.

[Yuli Ban]

[Yuli Ban]

[Yuli Ban]

He's not wrong

Artificial Intelligence is more profound than fire, electricity, or the internet, says Google boss

Artificial intelligence (AI ) and quantum computing will revolutionize our world over the next quarter-century predicted Sundar Pichai, the chief executive of Google’s owner Alphabet, in an interview.

The boss of the tech giant said AI, which is when machines are programmed to simulate the human mind and solve problems, will have a bigger impact than most of the major breakthroughs in recent history.

“I expect it to play a foundation role pretty much across every aspect of our lives,” he said, in a podcast recorded by BBC media editor Amol Rajan. “The progress in artificial intelligence is still in early stages but I view it as the most profound technology humanity will ever work on and we need to make sure we harness it to societies benefit.”

[weatheriscool]

Machine-learning algorithms used to detect Alzheimer's during phone conversations
https://medicalxpress.com/news/2021-07- ... tions.html
by Bob Yirka , Medical Xpress

Researchers working at the Department of Public Health, McCann Healthcare Worldwide Japan Inc., has created three algorithms that can be used to detect Alzheimer's in patients as they engage in phone conversations. The group has written a paper outlining the algorithms and their effectiveness and have uploaded it to the open-access site PLOS ONE.

Despite world-wide efforts, there still is no cure for Alzheimer's disease, which impacts millions of people around the globe including approximately 5.8 million in the United States. Medical researchers have made inroads towards slowing its progression, however; which is why it is becoming more and more important to identify the disease early. Thus scientists have turned their attention to finding new ways to predict which people will get the disease. In this new effort, the researchers have turned to machine learning as an aid to diagnoses.

Prior research has shown that some of the early signs of Alzheimer's include speaking more slowly than normal and pausing more often during conversations. Some work is already being done to recognize such speech difficulties—one project by a team in Japan uses the Telephone Interview for Cognitive Status (TICS-J) test, where phone conversations are recorded and studied to see if there is slow or broken speech. In this new study, the researchers have replaced the humans listening and analyzing phone conversations with a computer running a machine-learning algorithm.

[raklian]

[Yuli Ban]

Watch Robots Make Pizzas From Start to Finish at an Automated Pizzeria

The “show” starts with a robot grabbing a handful of dough and depositing it on a pan, where another bot flattens it, a third applies tomato sauce, etc. From dough-grabbing to inserting in the oven, preparing a pizza takes just 45 seconds. The oven can bake 6 pizzas at a time, yielding about 80 pizzas per hour. Once a pizza is baked to gooey perfection, a robot slices it and places it in a box, and it’s then transferred (by a robot, of course) to a numbered cubby from which the customer can retrieve it.

It’s a shame the pizzeria didn’t open during the height of the pandemic, as its revenues likely would have gone through the roof given that there’s zero person-to-person contact required for you to get a fresh, custom-made pizza in your hands (and more importantly, your belly!).

Pazzi’s creators spent eight years researching and developing the pizza bots, and they say the hardest part was getting the bots to work effectively with the raw dough. Since it’s made with yeast, the dough is sensitive to changes in temperature, humidity, and other factors, and for optimal results it needs to be rolled out and baked with very precise timing.