Tag Archives: rich
#437892 This Week’s Awesome Tech Stories From ...
ENVIRONMENT
Human-Made Stuff Now Outweighs All Life on Earth
Stephanie Pappas | Scientific American
“Humanity has reached a new milestone in its dominance of the planet: human-made objects may now outweigh all of the living beings on Earth. Roads, houses, shopping malls, fishing vessels, printer paper, coffee mugs, smartphones and all the other infrastructure of daily life now weigh in at approximately 1.1 trillion metric tons—equal to the combined dry weight of all plants, animals, fungi, bacteria, archaea and protists on the planet.”
SPACE
So, It Turns Out SpaceX Is Pretty Good at Rocketing
Eric Berger | Ars Technica
“As the Sun sank toward the South Texas horizon, a fantastical-looking spaceship rose into the reddening sky. It was, in a word, epic. …This was one heck of a test-flight that addressed a number of unknowns about Starship, which is the upper stage of SpaceX’s new launch system and may one day land humans on the Moon, Mars, and beyond.”
ARTIFICIAL INTELLIGENCE
Tiny Four-Bit Computers Are All You Need to Train AI
Karen Hao | MIT Technology Review
“The work…could increase the speed and cut the energy costs needed to train deep learning by more than sevenfold. It could also make training powerful AI models possible on smartphones and other small devices, which would improve privacy by helping to keep personal data on a local device. And it would make the process more accessible to researchers outside big, resource-rich tech companies.”
ENERGY
Did Quantum Scape Just Solve a 40-Year-Old Battery Problem?
Daniel Oberhaus | Wired
“[The properties of solid state batteries] would send…energy density through the roof, enable ultra-fast charging, and would eliminate the risk of battery fires. But for the past 40 years, no one has been able to make a solid-state battery that delivers on this promise—until earlier this year, when a secretive startup called QuantumScape claimed to have solved the problem. Now it has the data to prove it.”
ROBOTICS
Hyundai Buys Boston Dynamics for Nearly $1 Billion. Now What?
Evan Ackerman | IEEE Spectrum
“I hope that Boston Dynamics is unique enough that the kinds of rules that normally apply to robotics companies (or companies in general) can be set aside, at least somewhat, but I also worry that what made Boston Dynamics great was the explicit funding for the kinds of radical ideas that eventually resulted in robots like Atlas and Spot. Can Hyundai continue giving Boston Dynamics the support and freedom that they need to keep doing the kinds of things that have made them legendary? I certainly hope so.”
BIOTECH
CRISPR and Another Genetic Strategy Fix Cell Defects in Two Common Blood Disorders
Jocelyn Kaiser | Science
“It is a double milestone: new evidence that cures are possible for many people born with sickle cell disease and another serious blood disorder, beta-thalassemia, and a first for the genome editor CRISPR. Today, in The New England Journal of Medicine (NEJM) and tomorrow at the American Society of Hematology (ASH) meeting, teams report that two strategies for directly fixing malfunctioning blood cells have dramatically improved the health of a handful of people with these genetic diseases.”
ETHICS
The Dark Side of Big Tech’s Funding for AI Research
Tom Simonite | Wired
“Timnit Gebru’s exit from Google is a powerful reminder of how thoroughly companies dominate the field, with the biggest computers and the most resources. …[Meredith] Whittaker of AI Now says properly probing the societal effects of AI is fundamentally incompatible with corporate labs. ‘That kind of research that looks at the power and politics of AI is and must be inherently adversarial to the firms that are profiting from this technology.’i”
Image credit: Karsten Winegeart / Unsplash Continue reading →
Posted in Human Robots
#437564 How We Won the DARPA SubT Challenge: ...
This is a guest post. The views expressed here are those of the authors and do not necessarily represent positions of IEEE or its organizational units.
“Do you smell smoke?” It was three days before the qualification deadline for the Virtual Tunnel Circuit of the DARPA Subterranean Challenge Virtual Track, and our team was barrelling through last-minute updates to our robot controllers in a small conference room at the Michigan Tech Research Institute (MTRI) offices in Ann Arbor, Mich. That’s when we noticed the smell. We’d assumed that one of the benefits of entering a virtual disaster competition was that we wouldn’t be exposed to any actual disasters, but equipment in the basement of the building MTRI shares had started to smoke. We evacuated. The fire department showed up. And as soon as we could, the team went back into the building, hunkered down, and tried to make up for the unexpected loss of several critical hours.
Team BARCS joins the SubT Virtual Track
The smoke incident happened more than a year after we first learned of the DARPA Subterranean Challenge. DARPA announced SubT early in 2018, and at that time, we were interested in building internal collaborations on multi-agent autonomy problems, and SubT seemed like the perfect opportunity. Though a few of us had backgrounds in robotics, the majority of our team was new to the field. We knew that submitting a proposal as a largely non-traditional robotics team from an organization not known for research in robotics was a risk. However, the Virtual Track gave us the opportunity to focus on autonomy and multi-agent teaming strategies, areas requiring skill in asynchronous computing and sensor data processing that are strengths of our Institute. The prevalence of open source code, small inexpensive platforms, and customizable sensors has provided the opportunity for experts in fields other than robotics to apply novel approaches to robotics problems. This is precisely what makes the Virtual Track of SubT appealing to us, and since starting SubT, autonomy has developed into a significant research thrust for our Institute. Plus, robots are fun!
After many hours of research, discussion, and collaboration, we submitted our proposal early in 2018. And several months later, we found out that we had won a contract and became a funded team (Team BARCS) in the SubT Virtual Track. Now we needed to actually make our strategy work for the first SubT Tunnel Circuit competition, taking place in August of 2019.
Building a team of virtual robots
A natural approach to robotics competitions like SubT is to start with the question of “what can X-type robot do” and then build a team and strategy around individual capabilities. A particular challenge for the SubT Virtual Track is that we can’t design our own systems; instead, we have to choose from a predefined set of simulated robots and sensors that DARPA provides, based on the real robots used by Systems Track teams. Our approach is to look at what a team of robots can do together, determining experimentally what the best team configuration is for each environment. By the final competition, ideally we will be demonstrating the value of combining platforms across multiple Systems Track teams into a single Virtual Track team. Each of the robot configurations in the competition has an associated cost, and team size is constrained by a total cost. This provides another impetus for limiting dependence on complex sensor packages, though our ranging preference is 3D lidar, which is the most expensive sensor!
Image: Michigan Tech Research Institute
The teams can rely on realistic physics and sensors but they start off with no maps of any kind, so the focus is on developing autonomous exploratory behavior, navigation methods, and object recognition for their simulated robots.
One of the frequent questions we receive about the Virtual Track is if it’s like a video game. While it may look similar on the surface, everything under the hood in a video game is designed to service the game narrative and play experience, not require novel research in AI and autonomy. The purpose of simulations, on the other hand, is to include full physics and sensor models (including noise and errors) to provide a testbed for prototyping and developing solutions to those real-world challenges. We are starting with realistic physics and sensors but no maps of any kind, so the focus is on developing autonomous exploratory behavior, navigation methods, and object recognition for our simulated robots.
Though the simulation is more like real life than a video game, it is not real life. Due to occasional software bugs, there are still non-physical events, like the robots falling through an invisible hole in the world or driving through a rock instead of over it or flipping head over heels when driving over a tiny lip between world tiles. These glitches, while sometimes frustrating, still allow the SubT Virtual platform to be realistic enough to support rapid prototyping of controller modules that will transition straightforwardly onto hardware, closing the loop between simulation and real-world robots.
Full autonomy for DARPA-hard scenarios
The Virtual Track requirement that the robotic agents be fully autonomous, rather than have a human supervisor, is a significant distinction between the Systems and Virtual Tracks of SubT. Our solutions must be hardened against software faults caused by things like missing and bad data since our robots can’t turn to us for help. In order for a team of robots to complete this objective reliably with no human-in-the-loop, all of the internal systems, from perception to navigation to control to actuation to communications, must be able to autonomously identify and manage faults and failures anywhere in the control chain.
The communications limitations in subterranean environments (both real and virtual) mean that we need to keep the amount of information shared between robots low, while making the usability of that information for joint decision-making high. This goal has guided much of our design for autonomous navigation and joint search strategy for our team. For example, instead of sharing the full SLAM map of the environment, our agents only share a simplified graphical representation of the space, along with data about frontiers it has not yet explored, and are able to merge its information with the graphs generated by other agents. The merged graph can then be used for planning and navigation without having full knowledge of the detailed 3D map.
The Virtual Track requires that the robotic agents be fully autonomous. With no human-in-the-loop, all of the internal systems, from perception to navigation to control to actuation to communications, must be able to identify and manage faults and failures anywhere in the control chain.
Since the objective of the SubT program is to advance the state-of-the-art in rapid autonomous exploration and mapping of subterranean environments by robots, our first software design choices focused on the mapping task. The SubT virtual environments are sufficiently rich as to provide interesting problems in building so-called costmaps that accurately separate obstructions that are traversable (like ramps) from legitimately impassible obstructions. An extra complication we discovered in the first course, which took place in mining tunnels, was that the angle of the lowest beam of the lidar was parallel to the down ramps in the tunnel environment, so they could not “see” the ground (or sometimes even obstructions on the ramp) until they got close enough to the lip of the ramp to receive lidar reflections off the bottom of the ramp. In this case, we had to not only change the costmap to convince the robot that there was safe ground to reach over the lip of the ramp, but also had to change the path planner to get the robot to proceed with caution onto the top of the ramp in case there were previously unseen obstructions on the ramp.
In addition to navigation in the costmaps, the robot must be able to generate its own goals to navigate to. This is what produces exploratory behavior when there is no map to start with. SLAM is used to generate a detailed map of the environment explored by a single robot—the space it has probed with its sensors. From the sensor data, we are able to extract information about the interior space of the environment while looking for holes in the data, to determine things like whether the current tunnel continues or ends, or how many tunnels meet at an intersection. Once we have some understanding of the interior space, we can place navigation goals in that space. These goals naturally update as the robot traverses the tunnel, allowing the entire space to be explored.
Sending our robots into the virtual unknown
The solutions for the Virtual Track competitions are tested by DARPA in multiple sequestered runs across many environments for each Circuit in the month prior to the Systems Track competition. We must wait until the joint award ceremony at the conclusion of the Systems Track to find out the results, and we are completely in the dark about placings before the awards are announced. It’s nerve-wracking! The challenges of the worlds used in the Circuit events are also hand-designed, so features of the worlds we use for development could be combined in ways we have not anticipated—it’s always interesting to see what features were prioritized after the event. We test everything in our controllers well enough to feel confident that we at least are submitting something reasonably stable and broadly capable, and once the solution is in, we can’t really do anything other than “let go” and get back to work on the next phase of development. Maybe it’s somewhat like sending your kid to college: “we did our best to prepare you for this world, little bots. Go do good.”
Image: Michigan Tech Research Institute
The first SubT competition was the Tunnel Circuit, featuring a labyrinthine environment that simulated human-engineered tunnels, including hazards such as vertical shafts and rubble.
The first competition was the Tunnel Circuit, in October 2019. This environment models human-engineered tunnels. Two substantial challenges in this environment were vertical shafts and rubble. Our team accrued 21 points over 15 competition runs in five separate tunnel environments for a second place finish, behind Team Coordinated Robotics.
The next phase of the SubT virtual competition was the Urban Circuit. Much of the difference between our Tunnel and Urban Circuit results came down to thorough testing to identify failure modes and implementations of checks and data filtering for fault tolerance. For example, in the SLAM nodes run by a single robot, the coordinates of the most recent sensor data are changed multiple times during processing and integration into the current global 3D map of the “visited” environment stored by that robot. If there is lag in IMU or clock data, the observation may be temporarily registered at a default location that is very far from the actual position. Since most of our decision processes for exploration are downstream from SLAM, this can cause faulty or impossible goals to be generated, and the robots then spend inordinate amounts of time trying to drive through walls. We updated our method to add a check to see if the new map position has jumped a far distance from the prior map position, and if so, we threw that data out.
Image: Michigan Tech Research Institute
In open spaces like the rooms in the Urban circuit, we adjusted our approach to exploration through graph generation to allow the robots to accurately identify viable routes while helping to prevent forays off platform edges.
Our approach to exploration through graph generation based on identification of interior spaces allowed us to thoroughly explore the centers of rooms, although we did have to make some changes from the Tunnel circuit to achieve that. In the Tunnel circuit, we used a simplified graph of the environment based on landmarks like intersections. The advantage of this approach is that it is straightforward for two robots to compare how the graphs of the space they explored individually overlap. In open spaces like the rooms in the Urban circuit, we chose to instead use a more complex, less directly comparable graph structure based on the individual robot’s trajectory. This allowed the robots to accurately identify viable routes between features like subway station platforms and subway tracks, as well as to build up the navigation space for room interiors, while helping to prevent forays off the platform edges. Frontier information is also integrated into the graph, providing a uniform data structure for both goal selection and route planning.
The results are in!
The award ceremony for the Urban Circuit was held concurrently with the Systems Track competition awards this past February in Washington State. We sent a team representative to participate in the Technical Interchange Meeting and present the approach for our team, and the rest of us followed along from our office space on the DARPAtv live stream. While we were confident in our solution, we had also been tracking the online leaderboard and knew our competitors were going to be submitting strong solutions. Since the competition environments are hand-designed, there are always novel challenges that could be presented in these environments as well. We knew we would put up a good fight, but it was very exciting to see BARCS appear in first place!
Any time we implement a new module in our control system, there is a lot of parameter tuning that has to happen to produce reliably good autonomous behavior. In the Urban Circuit, we did not sufficiently test some parameter values in our exploration modules. The effect of this was that the robots only chose to go down small hallways after they explored everything else in their environment, which meant very often they ran out of time and missed a lot of small rooms. This may be the biggest source of lost points for us in the Urban Circuit. One of our major plans going forward from the Urban Circuit is to integrate more sophisticated node selection methods, which can help our robots more intelligently prioritize which frontier nodes to visit. By going through all three Circuit challenges, we will learn how to appropriately add weights to the frontiers based on features of the individual environments. For the Final Challenge, when all three Circuit environments will be combined into large systems, we plan to implement adaptive controllers that will identify their environments and use the appropriate optimized parameters for that environment. In this way, we expect our agents to be able to (for example) prioritize hallways and other small spaces in Urban environments, and perhaps prioritize large openings over small in the Cave environments, if the small openings end up being treacherous overall.
Next for our team: Cave Circuit
Coming up next for Team BARCS is the Virtual Cave Circuit. We are in the middle of testing our hypothesis that our controller will transition from UGVs to UAVs and developing strategies for refining our solution to handle Cave Circuit environmental hazards. The UAVs have a shorter battery life than the UGVs, so executing a joint exploration strategy will also be a high priority for this event, as will completing our work on graph sharing and merging, which will give our robot teams more sophisticated options for navigation and teamwork. We’re reaching a threshold in development where we can start increasing the “smarts” of the robots, which we anticipate will be critical for the final competition, where all of the challenges of SubT will be combined to push the limits of innovation. The Cave Circuit will also have new environmental challenges to tackle: dynamic features such as rock falls have been added, which will block previously accessible passages in the cave environment. We think our controllers are well-poised to handle this new challenge, and we’re eager to find out if that’s the case.
As of now, the biggest worries for us are time and team composition. The Cave Circuit deadline has been postponed to October 15 due to COVID-19 delays, with the award ceremony in mid-November, but there have also been several very compelling additions to the testbed that we would like to experiment with before submission, including droppable networking ‘breadcrumbs’ and new simulated platforms. There are design trade-offs when balancing general versus specialist approaches to the controllers for these robots—since we are adding UAVs to our team for the first time, there are new decisions that will have to be made. For example, the UAVs can ascend into vertical spaces, but only have a battery life of 20 minutes. The UGVs by contrast have 90 minute battery life. One of our strategies is to do an early return to base with one or more agents to buy down risk on making any artifact reports at all for the run, hedging against our other robots not making it back in time, a lesson learned from the Tunnel Circuit. Should a UAV take on this role, or is it better to have them explore deeper into the environment and instead report their artifacts to a UGV or network node, which comes with its own risks? Testing and experimentation to determine the best options takes time, which is always a worry when preparing for a competition! We also anticipate new competitors and stiffer competition all around.
Image: Michigan Tech Research Institute
Team BARCS has now a year to prepare for the final DARPA SubT Challenge event, expected to take place in late 2021.
Going forward from the Cave Circuit, we will have a year to prepare for the final DARPA SubT Challenge event, expected to take place in late 2021. What we are most excited about is increasing the level of intelligence of the agents in their teamwork and joint exploration of the environment. Since we will have (hopefully) built up robust approaches to handling each of the specific types of environments in the Tunnel, Urban, and Cave circuits, we will be aiming to push the limits on collaboration and efficiency among the agents in our team. We view this as a central research contribution of the Virtual Track to the Subterranean Challenge because intelligent, adaptive, multi-robot collaboration is an upcoming stage of development for integration of robots into our lives.
The Subterranean Challenge Virtual Track gives us a bridge for transitioning our more abstract research ideas and algorithms relevant to this degree of autonomy and collaboration onto physical systems, and exploring the tangible outcomes of implementing our work in the real world. And the next time there’s an incident in the basement of our building, the robots (and humans) of Team BARCS will be ready to respond.
Richard Chase, Ph.D., P.E., is a research scientist at Michigan Tech Research Institute (MTRI) and has 20 years of experience developing robotics and cyber physical systems in areas from remote sensing to autonomous vehicles. At MTRI, he works on a variety of topics such as swarm autonomy, human-swarm teaming, and autonomous vehicles. His research interests are the intersection of design, robotics, and embedded systems.
Sarah Kitchen is a Ph.D. mathematician working as a research scientist and an AI/Robotics focus area leader at MTRI. Her research interests include intelligent autonomous agents and multi-agent collaborative teams, as well as applications of autonomous robots to sensing systems.
This material is based upon work supported by the Defense Advanced Research Projects Agency (DARPA) under Contract No. HR001118C0124 and is released under Distribution Statement (Approved for Public Release, Distribution Unlimited). Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of DARPA. Continue reading →
Posted in Human Robots
#437407 Nvidia’s Arm Acquisition Brings the ...
Artificial intelligence and mobile computing have been two of the most disruptive technologies of this century. The unification of the two companies that made them possible could have wide-ranging consequences for the future of computing.
California-based Nvidia’s graphics processing units (GPUs) have powered the deep learning revolution ever since Google researchers discovered in 2011 that they could run neural networks far more efficiently than conventional CPUs. UK company Arm’s energy-efficient chip designs have dominated the mobile and embedded computing markets for even longer.
Now the two will join forces after the American company announced a $40 billion deal to buy Arm from its Japanese owner, Softbank. In a press release announcing the deal, Nvidia touted its potential to rapidly expand the reach of AI into all areas of our lives.
“In the years ahead, trillions of computers running AI will create a new internet-of-things that is thousands of times larger than today’s internet-of-people,” said Nvidia founder and CEO Jensen Huang. “Uniting NVIDIA’s AI computing capabilities with the vast ecosystem of Arm’s CPU, we can advance computing from the cloud, smartphones, PCs, self-driving cars and robotics, to edge IoT, and expand AI computing to every corner of the globe.”
There are good reasons to believe the hype. The two companies are absolutely dominant in their respective fields—Nvidia’s GPUs support more than 97 percent of AI computing infrastructure offered by big cloud service providers, and Arm’s chips power more than 90 percent of smartphones. And there’s little overlap in their competencies, which means the relationship could be a truly symbiotic one.
“I think the deal “fits like a glove” in that Arm plays in areas that Nvidia does not or isn’t that successful, while NVIDIA plays in many places Arm doesn’t or isn’t that successful,” analyst Patrick Moorhead wrote in Forbes.
One of the most obvious directions would be to expand Nvidia’s AI capabilities to the kind of low-power edge devices that Arm excels in. There’s growing demand for AI in devices like smartphones, wearables, cars, and drones, where transmitting data to the cloud for processing is undesirable either for reasons of privacy or speed.
But there might also be fruitful exchanges in the other direction. Huang told Moorhead a major focus would be bringing Arm’s expertise in energy efficiency to the data center. That’s a big concern for technology companies whose electricity bills and green credentials are taking a battering thanks to the huge amounts of energy required to run millions of computer chips around the clock.
The deal may not be plain sailing, though, most notably due to the two companies’ differing business models. While Nvidia sells ready-made processors, Arm simply creates chip designs and then licenses them to other companies who can then customize them to their particular hardware needs. It operates on an open-licence basis whereby any company with the necessary cash can access its designs.
As a result, its designs are found in products built by hundreds of companies that license its innovations, including Apple, Samsung, Huawei, Qualcomm, and even Nvidia. Some, including two of the company’s co-founders, have raised concerns that the purchase by Nvidia, which competes with many of these other companies, could harm the neutrality that has been central to its success.
It’s possible this could push more companies towards RISC-V, an open-source technology developed by researchers at the University of California at Berkeley that rivals Arm’s and is not owned by any one company. However, there are plenty of reasons why most companies still prefer arm over the less feature-rich open-source option, and it might take a considerable push to convince Arm’s customers to jump ship.
The deal will also have to navigate some thorny political issues. Unions, politicians, and business leaders in the UK have voiced concerns that it could lead to the loss of high-tech jobs, and government sources have suggested conditions could be placed on the deal.
Regulators in other countries could also put a spanner in the works. China is concerned that if Arm becomes US-owned, many of the Chinese companies that rely on its technology could become victims of export restrictions as the China-US trade war drags on. South Korea is also wary that the deal could create a new technology juggernaut that could dent Samsung’s growth in similar areas.
Nvidia has made commitments to keep Arm’s headquarters in the UK, which it says should lessen concerns around jobs and export restrictions. It’s also pledged to open a new world-class technology center in Cambridge and build a state-of-the-art AI supercomputer powered by Arm’s chips there. Whether the deal goes through still hangs in the balance, but of it does it could spur a whole new wave of AI innovation.
Image Credit: Nvidia Continue reading →
Posted in Human Robots
#436944 Is Digital Learning Still Second Best?
As Covid-19 continues to spread, the world has gone digital on an unprecedented scale. Tens of thousands of employees are working from home, and huge conferences, like the Google I/O and Apple WWDC software extravaganzas, plan to experiment with digital events.
Universities too are sending students home. This might have meant an extended break from school not too long ago. But no more. As lecture halls go empty, an experiment into digital learning at scale is ramping up. In the US alone, over 100 universities, from Harvard to Duke, are offering online classes to students to keep the semester going.
While digital learning has been improving for some time, Covid-19 may not only tip us further into a more digitally connected reality, but also help us better appreciate its benefits. This is important because historically, digital learning has been viewed as inferior to traditional learning. But that may be changing.
The Inversion
We often think about digital technologies as ways to reach people without access to traditional services—online learning for children who don’t have schools nearby or telemedicine for patients with no access to doctors. And while these solutions have helped millions of people, they’re often viewed as “second best” and “better than nothing.” Even in more resource-rich environments, there’s an assumption one should pay more to attend an event in person—a concert, a football game, an exercise class—while digital equivalents are extremely cheap or free. Why is this? And is the situation about to change?
Take the case of Dr. Sanjeev Arora, a professor of medicine at the University of New Mexico. Arora started Project Echo because he was frustrated by how many late-stage cases of hepatitis C he encountered in rural New Mexico. He realized that if he had reached patients sooner, he could have prevented needless deaths. The solution? Digital learning for local health workers.
Project Echo connects rural healthcare practitioners to specialists at top health centers by video. The approach is collaborative: Specialists share best practices and work through cases with participants to apply them in the real world and learn from edge cases. Added to expert presentations, there are lots of opportunities to ask questions and interact with specialists.
The method forms a digital loop of learning, practice, assessment, and adjustment.
Since 2003, Project Echo has scaled to 800 locations in 39 countries and trained over 90,000 healthcare providers. Most notably, a study in The New England Journal of Medicine found that the outcomes of hepatitis C treatment given by Project Echo trained healthcare workers in rural and underserved areas were similar to outcomes at university medical centers. That is, digital learning in this context was equivalent to high quality in-person learning.
If that is possible today, with simple tools, will they surpass traditional medical centers and schools in the future? Can digital learning more generally follow suit and have the same success? Perhaps. Going digital brings its own special toolset to the table too.
The Benefits of Digital
If you’re training people online, you can record the session to better understand their engagement levels—or even add artificial intelligence to analyze it in real time. Ahura AI, for example, founded by Bryan Talebi, aims to upskill workers through online training. Early study of their method suggests they can significantly speed up learning by analyzing users’ real-time emotions—like frustration or distraction—and adjusting the lesson plan or difficulty on the fly.
Other benefits of digital learning include the near-instantaneous download of course materials—rather than printing and shipping books—and being able to more easily report grades and other results, a requirement for many schools and social services organizations. And of course, as other digitized industries show, digital learning can grow and scale further at much lower costs.
To that last point, 360ed, a digital learning startup founded in 2016 by Hla Hla Win, now serves millions of children in Myanmar with augmented reality lesson plans. And Global Startup Ecosystem, founded by Christine Souffrant Ntim and Einstein Kofi Ntim in 2015, is the world’s first and largest digital accelerator program. Their entirely online programs support over 1,000 companies in 90 countries. It’s astonishing how fast both of these organizations have grown.
Notably, both examples include offline experiences too. Many of the 360ed lesson plans come with paper flashcards children use with their smartphones because the online-offline interaction improves learning. The Global Startup Ecosystem also hosts about 10 additional in-person tech summits around the world on various topics through a related initiative.
Looking further ahead, probably the most important benefit of online learning will be its potential to integrate with other digital systems in the workplace.
Imagine a medical center that has perfect information about every patient and treatment in real time and that this information is (anonymously and privately) centralized, analyzed, and shared with medical centers, research labs, pharmaceutical companies, clinical trials, policy makers, and medical students around the world. Just as self-driving cars can learn to drive better by having access to the experiences of other self-driving cars, so too can any group working to solve complex, time-sensitive challenges learn from and build on each other’s experiences.
Why This Matters
While in the long term the world will likely end up combining the best aspects of traditional and digital learning, it’s important in the near term to be more aware of the assumptions we make about digital technologies. Some of the most pioneering work in education, healthcare, and other industries may not be highly visible right now because it is in a virtual setting. Most people are unaware, for example, that the busiest emergency room in rural America is already virtual.
Once they start converging with other digital technologies, these innovations will likely become the mainstream system for all of us. Which raises more questions: What is the best business model for these virtual services? If they start delivering better healthcare and educational outcomes than traditional institutions, should they charge more? Hopefully, we will see an even bigger shift occurring, in which technology allows us to provide high quality education, healthcare, and other services to everyone at more affordable prices than today.
These are some of the topics we can consider as Covid-19 forces us into uncharted territory.
Image Credit: Andras Vas / Unsplash Continue reading →
Posted in Human Robots
#436482 50+ Reasons Our Favorite Emerging ...
For most of history, technology was about atoms, the manipulation of physical stuff to extend humankind’s reach. But in the last five or six decades, atoms have partnered with bits, the elemental “particles” of the digital world as we know it today. As computing has advanced at the accelerating pace described by Moore’s Law, technological progress has become increasingly digitized.
SpaceX lands and reuses rockets and self-driving cars do away with drivers thanks to automation, sensors, and software. Businesses find and hire talent from anywhere in the world, and for better and worse, a notable fraction of the world learns and socializes online. From the sequencing of DNA to artificial intelligence and from 3D printing to robotics, more and more new technologies are moving at a digital pace and quickly emerging to reshape the world around us.
In 2019, stories charting the advances of some of these digital technologies consistently made headlines. Below is, what is at best, an incomplete list of some of the big stories that caught our eye this year. With so much happening, it’s likely we’ve missed some notable headlines and advances—as well as some of your personal favorites. In either instance, share your thoughts and candidates for the biggest stories and breakthroughs on Facebook and Twitter.
With that said, let’s dive straight into the year.
Artificial Intelligence
No technology garnered as much attention as AI in 2019. With good reason. Intelligent computer systems are transitioning from research labs to everyday life. Healthcare, weather forecasting, business process automation, traffic congestion—you name it, and machine learning algorithms are likely beginning to work on it. Yet, AI has also been hyped up and overmarketed, and the latest round of AI technology, deep learning, is likely only one piece of the AI puzzle.
This year, Open AI’s game-playing algorithms beat some of the world’s best Dota 2 players, DeepMind notched impressive wins in Starcraft, and Carnegie Mellon University’s Libratus “crushed” pros at six-player Texas Hold‘em.
Speaking of games, AI’s mastery of the incredibly complex game of Go prompted a former world champion to quit, stating that AI ‘”cannot be defeated.”
But it isn’t just fun and games. Practical, powerful applications that make the best of AI’s pattern recognition abilities are on the way. Insilico Medicine, for example, used machine learning to help discover and design a new drug in just 46 days, and DeepMind is focused on using AI to crack protein folding.
Of course, AI can be a double-edged sword. When it comes to deepfakes and fake news, for example, AI makes both easier to create and detect, and early in the year, OpenAI created and announced a powerful AI text generator but delayed releasing it for fear of malicious use.
Recognizing AI’s power for good and ill, the OECD, EU, World Economic Forum, and China all took a stab at defining an ethical framework for the development and deployment of AI.
Computing Systems
Processors and chips kickstarted the digital boom and are still the bedrock of continued growth. While progress in traditional silicon-based chips continues, it’s slowing and getting more expensive. Some say we’re reaching the end of Moore’s Law. While that may be the case for traditional chips, specialized chips and entirely new kinds of computing are waiting in the wings.
In fall 2019, Google confirmed its quantum computer had achieved “quantum supremacy,” a term that means a quantum computer can perform a calculation a normal computer cannot. IBM pushed back on the claim, and it should be noted the calculation was highly specialized. But while it’s still early days, there does appear to be some real progress (and more to come).
Should quantum computing become truly practical, “the implications are staggering.” It could impact machine learning, medicine, chemistry, and materials science, just to name a few areas.
Specialized chips continue to take aim at machine learning—a giant new chip with over a trillion transistors, for example, may make machine learning algorithms significantly more efficient.
Cellular computers also saw advances in 2019 thanks to CRISPR. And the year witnessed the emergence of the first reprogrammable DNA computer and new chips inspired by the brain.
The development of hardware computing platforms is intrinsically linked to software. 2019 saw a continued move from big technology companies towards open sourcing (at least parts of) their software, potentially democratizing the use of advanced systems.
Networks
Increasing interconnectedness has, in many ways, defined the 21st century so far. Your phone is no longer just a phone. It’s access to the world’s population and accumulated knowledge—and it fits in your pocket. Pretty neat. This is all thanks to networks, which had some notable advances in 2019.
The biggest network development of the year may well be the arrival of the first 5G networks.
5G’s faster speeds promise advances across many emerging technologies.
Self-driving vehicles, for example, may become both smarter and safer thanks to 5G C-V2X networks. (Don’t worry with trying to remember that. If they catch on, they’ll hopefully get a better name.)
Wi-Fi may have heard the news and said “hold my beer,” as 2019 saw the introduction of Wi-Fi 6. Perhaps the most important upgrade, among others, is that Wi-Fi 6 ensures that the ever-growing number of network connected devices get higher data rates.
Networks also went to space in 2019, as SpaceX began launching its Starlink constellation of broadband satellites. In typical fashion, Elon Musk showed off the network’s ability to bounce data around the world by sending a Tweet.
Augmented Reality and Virtual Reality
Forget Pokemon Go (unless you want to add me as a friend in the game—in which case don’t forget Pokemon Go). 2019 saw AR and VR advance, even as Magic Leap, the most hyped of the lot, struggled to live up to outsized expectations and sell headsets.
Mixed reality AR and VR technologies, along with the explosive growth of sensor-based data about the world around us, is creating a one-to-one “Mirror World” of our physical reality—a digital world you can overlay on our own or dive into immersively thanks to AR and VR.
Facebook launched Replica, for example, which is a photorealistic virtual twin of the real world that, among other things, will help train AIs to better navigate their physical surroundings.
Our other senses (beyond eyes) may also become part of the Mirror World through the use of peripherals like a newly developed synthetic skin that aim to bring a sense of touch to VR.
AR and VR equipment is also becoming cheaper—with more producers entering the space—and more user-friendly. Instead of a wired headset requiring an expensive gaming PC, the new Oculus Quest is a wireless, self-contained step toward the mainstream.
Niche uses also continue to gain traction, from Google Glass’s Enterprise edition to the growth of AR and VR in professional education—including on-the-job-training and roleplaying emotionally difficult work encounters, like firing an employee.
Digital Biology and Biotech
The digitization of biology is happening at an incredible rate. With wild new research coming to light every year and just about every tech giant pouring money into new solutions and startups, we’re likely to see amazing advances in 2020 added to those we saw in 2019.
None were, perhaps, more visible than the success of protein-rich, plant-based substitutes for various meats. This was the year Beyond Meat was the top IPO on the NASDAQ stock exchange and people stood in line for the plant-based Impossible Whopper and KFC’s Beyond Chicken.
In the healthcare space, a report about three people with HIV who became virus free thanks to a bone marrow transplants of stem cells caused a huge stir. The research is still in relatively early stages, and isn’t suitable for most people, but it does provides a glimmer of hope.
CRISPR technology, which almost deserves its own section, progressed by leaps and bounds. One tweak made CRISPR up to 50 times more accurate, while the latest new CRISPR-based system, CRISPR prime, was described as a “word processor” for gene editing.
Many areas of healthcare stand to gain from CRISPR. For instance, cancer treatment, were a first safety test showed ‘promising’ results.
CRISPR’s many potential uses, however, also include some weird/morally questionable areas, which was exemplified by one the year’s stranger CRISPR-related stories about a human-monkey hybrid embryo in China.
Incidentally, China could be poised to take the lead on CRISPR thanks to massive investments and research programs.
As a consequence of quick advances in gene editing, we are approaching a point where we will be able to design our own biology—but first we need to have a serious conversation as a society about the ethics of gene editing and what lines should be drawn.
3D Printing
3D printing has quietly been growing both market size and the objects the printers are capable of producing. While both are impressive, perhaps the biggest story of 2019 is their increased speed.
One example was a boat that was printed in just three days, which also set three new world records for 3D printing.
3D printing is also spreading in the construction industry. In Mexico, the technology is being used to construct 50 new homes with subsidized mortgages of just $20/month.
3D printers also took care of all parts of a 640 square-meter home in Dubai.
Generally speaking, the use of 3D printing to make parts for everything from rocket engines (even entire rockets) to trains to cars illustrates the sturdiness of the technology, anno 2019.
In healthcare, 3D printing is also advancing the cause of bio-printed organs and, in one example, was used to print vascularized parts of a human heart.
Robotics
Living in Japan, I get to see Pepper, Aibo, and other robots on pretty much a daily basis. The novelty of that experience is spreading to other countries, and robots are becoming a more visible addition to both our professional and private lives.
We can’t talk about robots and 2019 without mentioning Boston Dynamics’ Spot robot, which went on sale for the general public.
Meanwhile, Google, Boston Dynamics’ former owner, rebooted their robotics division with a more down-to-earth focus on everyday uses they hope to commercialize.
SoftBank’s Pepper robot is working as a concierge and receptionist in various countries. It is also being used as a home companion. Not satisfied, Pepper rounded off 2019 by heading to the gym—to coach runners.
Indeed, there’s a growing list of sports where robots perform as well—or better—than humans.
2019 also saw robots launch an assault on the kitchen, including the likes of Samsung’s robot chef, and invade the front yard, with iRobot’s Terra robotic lawnmower.
In the borderlands of robotics, full-body robotic exoskeletons got a bit more practical, as the (by all accounts) user-friendly, battery-powered Sarcos Robotics Guardian XO went commercial.
Autonomous Vehicles
Self-driving cars did not—if you will forgive the play on words—stay quite on track during 2019. The fallout from Uber’s 2018 fatal crash marred part of the year, while some big players ratcheted back expectations on a quick shift to the driverless future. Still, self-driving cars, trucks, and other autonomous systems did make progress this year.
Winner of my unofficial award for best name in self-driving goes to Optimus Ride. The company also illustrates that self-driving may not be about creating a one-size-fits-all solution but catering to specific markets.
Self-driving trucks had a good year, with tests across many countries and states. One of the year’s odder stories was a self-driving truck traversing the US with a delivery of butter.
A step above the competition may be the future slogan (or perhaps not) of Boeing’s self-piloted air taxi that saw its maiden test flight in 2019. It joins a growing list of companies looking to create autonomous, flying passenger vehicles.
2019 was also the year where companies seemed to go all in on last-mile autonomous vehicles. Who wins that particular competition could well emerge during 2020.
Blockchain and Digital Currencies
Bitcoin continues to be the cryptocurrency equivalent of a rollercoaster, but the underlying blockchain technology is progressing more steadily. Together, they may turn parts of our financial systems cashless and digital—though how and when remains a slightly open question.
One indication of this was Facebook’s hugely controversial announcement of Libra, its proposed cryptocurrency. The company faced immediate pushback and saw a host of partners jump ship. Still, it brought the tech into mainstream conversations as never before and is putting the pressure on governments and central banks to explore their own digital currencies.
Deloitte’s in-depth survey of the state of blockchain highlighted how the technology has moved from fintech into just about any industry you can think of.
One of the biggest issues facing the spread of many digital currencies—Bitcoin in particular, you could argue—is how much energy it consumes to mine them. 2019 saw the emergence of several new digital currencies with a much smaller energy footprint.
2019 was also a year where we saw a new kind of digital currency, stablecoins, rise to prominence. As the name indicates, stablecoins are a group of digital currencies whose price fluctuations are more stable than the likes of Bitcoin.
In a geopolitical sense, 2019 was a year of China playing catch-up. Having initially banned blockchain, the country turned 180 degrees and announced that it was “quite close” to releasing a digital currency and a wave of blockchain-programs.
Renewable Energy and Energy Storage
While not every government on the planet seems to be a fan of renewable energy, it keeps on outperforming fossil fuel after fossil fuel in places well suited to it—even without support from some of said governments.
One of the reasons for renewable energy’s continued growth is that energy efficiency levels keep on improving.
As a result, an increased number of coal plants are being forced to close due to an inability to compete, and the UK went coal-free for a record two weeks.
We are also seeing more and more financial institutions refusing to fund fossil fuel projects. One such example is the European Investment Bank.
Renewable energy’s advance is tied at the hip to the rise of energy storage, which also had a breakout 2019, in part thanks to investments from the likes of Bill Gates.
The size and capabilities of energy storage also grew in 2019. The best illustration came from Australia were Tesla’s mega-battery proved that energy storage has reached a stage where it can prop up entire energy grids.
Image Credit: Mathew Schwartz / Unsplash Continue reading →
Posted in Human Robots