Our Strength Lies in Our Humanity

Author Archives: Syndicated

Not Bot, Not Beast: Scientists Create First Ever Living, Programmable Organism

· · Posted in General, Robotics, Science and Society, Topics ·

Not Bot, Not Beast: Scientists Create First Ever Living, Programmable Organism

A remarkable combination of artificial intelligence (AI) and biology has produced the world’s first “living robots.”

This week, a research team of roboticists and scientists published their recipe for making a new lifeform called xenobots from stem cells. The term “xeno” comes from the frog cells (Xenopus laevis) used to make them.

One of the researchers described the creation as “neither a traditional robot nor a known species of animal,” but a “new class of artifact: a living, programmable organism.”

Xenobots are less than 1 millimeter long and made of 500-1,000 living cells. They have various simple shapes, including some with squat “legs.” They can propel themselves in linear or circular directions, join together to act collectively, and move small objects. Using their own cellular energy, they can live up to 10 days.

While these “reconfigurable biomachines” could vastly improve human, animal, and environmental health, they raise legal and ethical concerns.

Strange New ‘Creature’

To make xenobots, the research team used a supercomputer to test thousands of random designs of simple living things that could perform certain tasks.

The computer was programmed with an AI “evolutionary algorithm” to predict which organisms would likely display useful tasks, such as moving towards a target.

After the selection of the most promising designs, the scientists attempted to replicate the virtual models with frog skin or heart cells, which were manually joined using microsurgery tools. The heart cells in these bespoke assemblies contract and relax, giving the organisms motion.

The creation of xenobots is groundbreaking. Despite being described as “programmable living robots,” they are actually completely organic and made of living tissue. The term “robot” has been used because xenobots can be configured into different forms and shapes, and “programmed” to target certain objects, which they then unwittingly seek. They can also repair themselves after being damaged.

Possible Applications

Xenobots may have great value. Some speculate they could be used to clean our polluted oceans by collecting microplastics. Similarly, they may be used to enter confined or dangerous areas to scavenge toxins or radioactive materials. Xenobots designed with carefully shaped “pouches” might be able to carry drugs into human bodies.

Future versions may be built from a patient’s own cells to repair tissue or target cancers. Being biodegradable, xenobots would have an edge on technologies made of plastic or metal.

Further development of biological “robots” could accelerate our understanding of living and robotic systems. Life is incredibly complex, so manipulating living things could reveal some of life’s mysteries—and improve our use of AI.

Legal and Ethical Questions

Conversely, xenobots raise legal and ethical concerns. In the same way they could help target cancers, they could also be used to hijack life functions for malevolent purposes.

Some argue artificially making living things is unnatural, hubristic, or involves “playing God.” A more compelling concern is that of unintended or malicious use, as we have seen with technologies in fields including nuclear physics, chemistry, biology and AI. For instance, xenobots might be used for hostile biological purposes prohibited under international law.

More advanced future xenobots, especially ones that live longer and reproduce, could potentially “malfunction” and go rogue, and out-compete other species.

For complex tasks, xenobots may need sensory and nervous systems, possibly resulting in their sentience. A sentient programmed organism would raise additional ethical questions. Last year, the revival of a disembodied pig brain elicited concerns about different species’ suffering.

Managing Risks

The xenobot’s creators have rightly acknowledged the need for discussion around the ethics of their creation. The 2018 scandal over using CRISPR (which allows the introduction of genes into an organism) may provide an instructive lesson here. While the experiment’s goal was to reduce the susceptibility of twin baby girls to HIV-AIDS, associated risks caused ethical dismay. The scientist in question is in prison.

When CRISPR became widely available, some experts called for a moratorium on heritable genome editing. Others argued the benefits outweighed the risks.

While each new technology should be considered impartially and based on its merits, giving life to xenobots raises certain significant questions:

  1. Should xenobots have biological kill-switches in case they go rogue?
  2. Who should decide who can access and control them?
  3. What if “homemade” xenobots become possible? Should there be a moratorium until regulatory frameworks are established? How much regulation is required?

Lessons learned in the past from advances in other areas of science could help manage future risks, while reaping the possible benefits.

Long Road Here, Long Road Ahead

The creation of xenobots had various biological and robotic precedents. Genetic engineering has created genetically modified mice that become fluorescent in UV light.

Designer microbes can produce drugs and food ingredients that may eventually replace animal agriculture. In 2012, scientists created an artificial jellyfish called a “medusoid” from rat cells.

Robotics is also flourishing. Nanobots can monitor people’s blood sugar levels and may eventually be able to clear clogged arteries. Robots can incorporate living matter, which we witnessed when engineers and biologists created a sting-ray robot powered by light-activated cells.

In the coming years, we are sure to see more creations like xenobots that evoke both wonder and due concern. And when we do, it is important we remain both open-minded and critical.The Conversation

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Image Credit: Photo by Joel Filipe on Unsplash

Source: Singularity Hub:  https://singularityhub.com/2020/01/22/not-bot-not-beast-scientists-create-first-ever-living-programmable-organism/

The Brain Predicts Reward Like an AI, Says New DeepMind Research

· · Posted in General, Neuroscience, Science and Society, Topics ·

The Brain Predicts Reward Like an AI, Says New DeepMind Research

The idea of reinforcement learning—or learning based on reward—has been around for so long it’s easy to forget we don’t really know how it works.

If DeepMind’s new bombshell paper in Nature is any indication, a common approach in AI, one that’s led to humanity’s defeat in the game of Go against machines, may have the answer.

We all subconsciously learn complex behaviors in response to positive and negative feedback, but how that works in the brain remains a century-long mystery. By examining a powerful variant of reinforcement learning, dubbed distributional reinforcement learning, that outperforms original methods, the team suggests that the brain may simultaneously represent multiple predicted futures in parallel. Each future is assigned a different probability, or chance of actually occurring, based on reward.

Here’s the kicker: the team didn’t leave it as an AI-inspired hypothesis. In a collaboration with a lab at Harvard University, they recording straight from a mouse’s brain, and found signs of their idea encoded in its reward-processing neurons.

“Neuroscience and AI have always inspired one another. Reinforcement learning is one of the central approaches in contemporary AI research and was originally motivated by thinking about how rewards could be used to reinforce certain behaviors,” said Dr. Demis Hassabis, co-founder and CEO of DeepMind. “In this new work, we find that distributional reinforcement learning…may also be at work in human biology.”

Dopamine Rush

You’ve heard of “reward-based learning.” Take Pavlov’s dogs, a famous experiment in the early 1900s. Russian physiologist Ivan Pavlov repeatedly rang a bell before feeding his dogs, and found the dogs learned to instinctively salivate as soon as they heard the bell, no food in sight. They learned that the bell—something not associated with food at all—predicted food, the reward.

Neuroscientists eventually figured out that dopamine, a chemical messenger active in reward circuits in the brain, is involved in processing reward signals. Although dopaminergic neurons—neurons that release dopamine—are often associated with a “rush” or “high,” that’s not to say they make us “happy” or “feel good” per se.

Rather, dopaminergic neurons are high-rollers at a betting game. They constantly make predictions about the chance of receiving a reward, and only change how much dopamine they release if the prediction is off. A spot-on bet—mom calls you to lunch, you get lunch—and the neurons stay quiet. If the reward is better or larger than predicted, however, they amp up their activity and shoot off packets of dopamine; if the result is worse or less than expected, the neurons lower their activity. With time, dopaminergic neurons learn to adjust their predictions to better match the real world. In other words, dopamine doesn’t equate to “happiness” or “high”; rather, it responds to predictions’ errors. We learn by trying to correct that error.

These ideas aren’t just hypotheses. Using electrodes and chemical probes inside living brains, neuroscientists can watch dopaminergic neurons in action as animals are challenged with tasks that eventually lead to rewards—food, water, sex, and drugs.

But a question remained: do dopaminergic neurons work as a unit, encoding for a single expected outcome? Or do they diverge in their predictions?

Meanwhile in AI Land…

While neuroscientists busily examined living brains, AI researchers directly tested their high-level ideas in machines.

Enter reinforcement learning. When challenged with a task, an AI algorithm starts out with random predictions. It then takes an action, observes whether it gets a reward, and adjusts its predictions based on reality. After millions of trials, the AI hopefully minimizes its prediction errors, meaning that it knows exactly how to solve the task. Step by step, it can then tackle extremely complex problems, such as beating a human champion at Go.

That’s the standard narrative. When AI researchers dug into the weeds, however, they immediately faced a head-scratcher: how do you represent the chance for reward? The traditional approach is to give it an average number—a general “gut feeling,” so to speak, based on classic reward-learning theory in neuroscience. But in the real world, the chance of getting a reward is never perfectly averaged into a fixed number.

“When playing the lottery, for example, people expect to either win big, or win nothing—no one is thinking about getting the average outcome,” said first author Dr. Will Dabney, a research scientist at DeepMind.

In 2017, DeepMind researchers decided to encode that randomness into reinforcement learning. Rather than affixing a single number for prediction error, they swapped it for a sophisticated distribution of probabilities. In distributional reinforcement learning, the AI algorithm predicts a full spectrum of future rewards: some are more optimistic and amplify their reward signals when the reward is larger than expected; others more pessimistic, lowering their reward signals when it’s smaller than predicted.

Sound familiar? The ties to the brain’s reward-based predictive powers are hard to ignore. This led the DeepMind team to ask: can we see a similar distribution in individual dopaminergic neural predictions as well?

Come Together

Partnering with Harvard, the teams tested out their idea in the brains of mice.

The mice were taught to work for a juice reward, but how much tasty juice—a drop or a gush—was highly unpredictable, basically relying on the roll of a seven-sided die. As the mice struggled to predict their next sugary “hit,” the team watched the electrical activity of individual dopaminergic neurons in their brains’ reward regions.

In contrast to neuroscience canon, the team said, dopaminergic neurons didn’t act as one. Rather than collectively encoding for a single expected outcome, they were each “tuned” to a different prediction, with some expecting a larger amount of juice reward, and others less hopeful, predicting smaller volumes. When the team mapped out the whole swath of predictions, the distribution closely matched reality—the distribution of actual rewards.

“We found that dopaminergic neurons in the brain were each tuned to different levels of pessimism or optimism. If they were a choir, they wouldn’t all be singing the same note, but harmonizing,” the authors wrote.

In other words, they seemed to operate on very similar principles to distributed reinforcement learning, a powerful method in AI.

That’s great news for both AI and neuroscience. Although AI researchers empirically already know that distributional reinforcement learning, combined with deep neural networks, is extremely powerful, the new data further validates it as a potential path towards AI that learns in a manner more similar to human brains—the only example of high-level intelligence we know of.

“When we’re able to demonstrate that the brain employs algorithms like those we are using in our AI work, it bolsters our confidence that those algorithms will be useful in the long run—that they will scale well to complex real-world problems, and interface well with other computational processes. There’s a kind of validation involved: if the brain is doing it, it’s probably a good idea,” said senior author Dr. Matt Botvinick, Director of Neuroscience Research at DeepMind.

But neuroscience also has much to gain. If the brain follows a similar reward “program,” what happens if it preferentially listens to optimistic rather than pessimistic dopaminergic neurons? Does it cause impulsivity or depression? What mechanisms in the brain’s reward circuits, or single dopaminergic neurons, can shift reward prediction? Do these changes lead to drug addiction, gambling, and other unhealthy behaviors? And finally, by borrowing “back” ideas from AI, can we finally understand what motivates us, why we take risks, and what makes us tick?

“As the fields of AI and neuroscience try to understand the nature of intelligence, we continually borrow methods and inspiration from each other. I feel extremely lucky to have been able to add to this ongoing conversation, as we try to piece together the fundamental algorithms behind learning and intelligence, be it biological or artificial,” said Dabney.

Image Credit: Image by Raman Oza from Pixabay

Source: Singularity Hub:  https://singularityhub.com/2020/01/21/the-brain-predicts-reward-like-an-ai-says-new-deepmind-research/

World’s Largest Offshore Wind Farm Will Power 4.5 Million Homes

· · Posted in Creative Commons, Energy, Future of Energy, General, Science and Society, Topics ·

World’s Largest Offshore Wind Farm Will Power 4.5 Million Homes

Renewable energy statistics just keep topping each other. Solar power is getting cheaper. Battery storage capacity is getting better. And wind farms are getting bigger.

2019 saw the world’s biggest (at the time) offshore wind farm come online, as well as construction of the biggest offshore wind farm in the US off the coast of Atlantic City.

But a new figure blows all of these out of the water. Last week, British renewable energy developer SSE announced construction of Dogger Bank Wind Farm off the eastern coast of England in the North Sea.

With a capacity of 3.6 gigawatts (GW), Dogger Bank will be three times bigger than the world’s biggest existing wind farm, the nearby 1.2 GW Hornsea One.

Located near a seaside town called Ulrome, which is 195 miles north of London, Dogger Bank will have three separate sites—Creyke Beck A, Creyke Beck B, and Teesside A—each with a 1.2 GW capacity, and construction is slated to take two years.

The project is a collaboration between SSE and Equinor, a Norwegian energy company.

Hornsea One, which is still under construction and estimated to be done later this year, will power up to a million homes. That seems small in comparison to the figure SSE has estimated for Dogger Bank: its three sites together will allegedly be able to power up to 4.5 million homes.

It’s important to note that 4.5 million U.K. homes is equivalent to far fewer homes in the US, as we Americans use embarrassingly more energy than the British; our annual average electricity consumption per household in 2018 was 10,972 kWh—almost triple the U.K.’s 2017 household average of 3,760 kWh. Makes you wonder if we couldn’t scale back on our giant appliances, around-the-clock air conditioning, and ever-ready hot water.

Besides being the biggest wind farm in the world (unless an even bigger one is announced before it’s done? At the rate these things are popping up, that wouldn’t be surprising), Dogger Bank will also be one of the first to use the world’s biggest turbine. GE Renewable Energy’s Haliade-X was tested in the Port of Rotterdam last summer; it’s 260 meters (853 feet) tall (almost the same height as the GE building in Rockefeller Center) and its blades are 107 meters long (351 feet, just a tad shorter than an American football field).

Once those blades are twirling in the North Sea’s winds (and twirl they will—wind speeds in this area average 16-22 miles per hour; the top 3 windiest cities in the US seem breezy in comparison with speeds from 13.1-13.9 mph), each Haliade-X will be able to produce 12 megawatts of power.

Offshore wind has taken off over the past few years, and will continue to grow as the technology it requires becomes steadily better, cheaper, and more efficient. The new turbines being built now already generate three times as much energy as turbines from just five years ago; that’s an impressive pace, and means the expense of building wind farms are much more worth companies’ while.

They’re worth our while, too. Given the ecological and climate disasters increasingly filling our news feeds, it’s safe to say we can’t increase our renewable energy development fast enough.

Image Credit: Image courtesy of GE Renewable Energy

Source: Singularity Hub:  https://singularityhub.com/2020/01/20/worlds-largest-offshore-wind-farm-will-power-4-5-million-homes/

Precision Fermentation: What It Is, and How It Could Make Farming Obsolete

· · Posted in Future of Food, General, Science and Society, Topics ·

Precision Fermentation: What It Is, and How It Could Make Farming Obsolete

Concerns around the environmental impact of modern farming are causing us to reassess our relationship with food. But are claims that lab-grown food will replace agriculture in the coming decades realistic?

British journalist and environmental activist George Monbiot made a stir last week with the release of his documentary Apocalypse Cow, in which he claims that a technology called “precision fermentation” could make traditional farming obsolete in just a few decades.

In an article for The Guardian he contends that we will be able to produce the vast majority of our nutrients through industrial-scale brewing of single-celled organisms specially designed to produce particular compounds.

This idea of lab-grown food is certainly in the ascendancy at the minute. Concerns about the ethics and environmental impact of meat-eating are leading to growing numbers of people adopting vegetarianism and veganism, but they’ve also spawned a host of startups promising to grow “cultured meat” in the lab without ever involving real animals.

But Monbiot goes further, claiming the same technology could also replace arable farming and would be orders of magnitude more energy, water, and space-efficient than current farming practices, which could reverse agriculture’s hugely damaging effect on the environment.

Monbiot isn’t alone in predicting how disruptive precision fermentation could be. A recent report from think tank RethinkX predicted that lab-grown protein will be 10 times cheaper than animal protein by 2035, which will see the US beef and dairy industries’ production fall 50 percent by 2030, effectively bankrupting them.

Others are more skeptical, pointing to the huge practical and cultural barriers to such change—not least, what to do with the 28 percent of the planet that relies on agriculture for employment. Some of these criticisms are ideological ripostes to Monbiot’s equally ideological framing of the case, while farming lobbies have inevitably had their say. But others have highlighted some of the holes in the logic behind his claims.

In making his case for lab-grown food, Monbiot highlights the technology of Solar Foods, which uses solar energy to split water into oxygen and hydrogen to provide energy for microbes that then produce feedstocks. He claims this process is ten times as efficient as photosynthesis, and because everything is done in vats the land efficiency is roughly 20,000 times greater than arable farming.

But the company clarified to New Scientist that when you take into account the land required for solar panels, it would only be around ten times more efficient than farming soya. More importantly, renewable energy like solar is still a limited resource at present, and one that can only be expanded so fast.

Diverting it to power agriculture would inevitably lead to boosting carbon emissions elsewhere. And research has already shown that cultured meat produced using less sustainable energy sources could actually be worse for the climate.

Monbiot’s plans also assume huge improvements in the capabilities of lab-grown food companies to create food that is both nutritious and palatable. Cultured meat is still a long way from being able to recreate the combinations of muscle, fat, and connective tissue that make a steak so delicious.

More importantly, it will be incredibly hard to engineer anything as efficient as a plant at using solar energy to create food with the perfect balance of vitamins, minerals, and fiber, Cambridge University plant scientist David Hanke writes in a letter responding to Monbiot’s Guardian article.

While supplying minerals to microbes would likely rely on intensive chemical processes, plants are capable of extracting trace elements from the environment. They can also build solar harvesting equipment without the huge cost and pollution that would be involved in building enough solar panels to power agriculture.

All of this is not to say there isn’t a place for lab-grown food. Current agricultural practices are in many cases unsustainable, and using this kind of technology to replace the worst offenders is probably a smart idea. But at best it’s going to be complimentary to existing agricultural practices, and unlikely to be adopted in developing countries, where the capital expenditure involved will make it prohibitive.

There are also far more pressing and tractable places to apply technology if you want to improve the efficiency of our food supply chains, and therefore cut down their costs and environmental impact. The great tragedy of modern agriculture is that we already produce enough food for ten billion people, but a third of it goes to waste.

Better cold chain and smarter distribution systems could take a sizable chunk out of that, so maybe that would be smarter place to start.

Image Credit: Image by Free-Photos from Pixabay

Source: Singularity Hub:  https://singularityhub.com/2020/01/19/precision-fermentation-what-it-is-and-how-it-could-make-farming-obsolete/

This Week’s Awesome Tech Stories From Around the Web (Through January 18)

· · Posted in Curation, General, Science and Society, Topics ·

This Week’s Awesome Tech Stories From Around the Web (Through January 18)

BIOTECH

Meet Xenobot, an Eerie New Kind of Programmable Organism
Matt Simon | Wired
“A xenobot is a one-of-a-kind organism: It’s both a living thing made of living cells and a machine that the researchers can program to express certain behaviors. The frog cells aren’t special in and of themselves—it’s the emergent behavior they collectively produce that’s so remarkable.”

AUGMENTED REALITY

The Making of Mojo, Augmented Reality Contact Lenses That Give Your Eyes Superpowers
Mark Sullivan | Fast Company
“This audacious idea is part of a much larger trend. In the coming decade, it’s likely that our computing devices will become more personal and reside closer to—or even inside—our bodies. Our eyes are the logical next stop on the journey.”

ROBOTICS

Scientists Built a Robot From 40 Pigeon Feathers and It Flies Beautifully
Ryan F. Mandelbaum | Gizmodo
“This new study on pigeon wings has not only provided a simpler model for how bird wings work but allowed engineers to integrate that knowledge into a nimble flying machine. The researchers hope that PigeonBot will provide an inspiration for those constructing flying machines, as well as those studying birds.”

TECHNOLOGY

Google Reaches $1 Trillion Value, Even as It Faces New Tests
Daisuke Wakabayashi | The New York Times
“On Thursday, Google hit another eye-popping number. The market cap of Alphabet vaulted above $1 trillion for the first time. That made it the fourth technology company—after Apple, Amazon, and Microsoft over the past two years—to pass this once unimaginable valuation.”

FUTURE

The Case For…Cities That Aren’t Dystopian Surveillance States
Cory Doctorow | The Guardian
“Imagine a human-centered smart city that knows everything it can about things. It knows how many seats are free on every bus, it knows how busy every road is, it knows where there are short-hire bikes available and where there are potholes. …What it doesn’t know is anything about individuals in the city.”

Image Credit: Alexander Sinn / Unsplash

Source: Singularity Hub:  https://singularityhub.com/2020/01/18/this-weeks-awesome-tech-stories-from-around-the-web-through-january-18/

Mojo Vision’s Augmented Reality Contact Lenses Kick off a Race to AR on Your Eye

· · Posted in Creative Commons, General, Science and Society, Topics, Virtual Reality ·

Mojo Vision’s Augmented Reality Contact Lenses Kick off a Race to AR on Your Eye

The digital world has been creeping closer to your face. Was a time when a laptop was about as personal as you got with a computer. Then came smartphones, and a few years later, some of us started strapping screens to our heads. Now, in what must be the penultimate step to tapping brains directly, the world has a working digital display in a contact lens.

After years of quiet research, startup Mojo Vision demonstrated their augmented reality contact lens to select journalists at CES and then, yesterday, broke the news officially to a flurry of articles and first-hand accounts.

The Digital World Straight to Your Eyeballs

The research origins of the Mojo Lens date back to 2008. But even when the company was founded in 2015—with engineers from Google, Amazon, Microsoft, and a list of other big tech companies—much of the technology still had to be “called in from the future.”

“We’ve had to invent most everything we put in the lens,” Steven Sinclair, vice president of product and marketing at Mojo told VentureBeat. “As you can imagine, we’ve invented our own display. We’ve invented our own oxygenation system, we’ve invented our own power data, we’ve invented our own ASICS (custom chips) and power management tools. We’re inventing our own algorithms for eye-tracking.”

Though the Mojo Lens is very much still a prototype, some of the details are pretty mind blowing. The MicroLED display, which sits slightly off the surface of the eye at the center of the lens, is perhaps its most marvelous feature. There are no fewer than 70,000 pixels packed into an area the size of a grain of sand. That’s 14K resolution (for those of you with a paltry 4K TV).

Revealed previously, the display was the first clue Mojo was working on something special. There was speculation back then that the company, which said it was building an AR platform, was actually working on a contact lens. But Mojo would neither confirm nor deny it.

The lens itself isn’t the soft, disposable kind. It’s a hard scleral lens and will have to be used over and over, which means it’ll sit wirelessly charging in a bath of disinfectant each night. The lenses would also have to be replaced annually, according to the company (no word on cost yet).

Currently, the display is the primary piece of technology being demoed, but they’re already working to embed other electronics, like image and motion sensors, radios, and batteries. A product development roadmap shows the lenses gaining more and more electronics, version to version, with a final product including a painted iris to camouflage the lot.

At CES, the company gave journalists two demos, one of the contact lens itself (though no one was allowed to put it in their eye, for obvious reasons) and one of a future operating system (in VR only, for now).

The former showed a simple, green Google-Glass-like display that scrolled pre-loaded information, such as a text from a friend and the weather. (One pretty useful app might be a digital teleprompter for your eye to aid public speaking.) Then, in a dimly lit room, the prototype used edge-detection to outline traffic signs and people. Meanwhile, the operating system showed off future functionality, such as eye-tracking, to allow users to navigate a simple menu on the periphery by looking at an arrow icon that then enlarges information.

The end goal is to make Mojo Lens a pair of basically invisible, high-fidelity, wireless displays that deliver the digital world straight to your eyeballs, and the company has raised over $100 million to make it happen. At present, however, there’s a lot more work to be done.

This Is Not For You

We all still remember Google Glass and Magic Leap. At least, Mojo and many tech writers do, because the press campaign for the prototype lens was relatively restrained and the subsequent coverage has been suitably amazed but also politely (or outright) skeptical.

Though based on cool new technologies with plenty of potential, early AR headsets that overpromised and aimed at the mainstream have since ratcheted back expectations. The future vision is no less compelling, but the reality is that it takes time to refine designs and find worthwhile applications.

“We didn’t want to get overhyped and show something that was just vaporware,” Sinclair told CNET.

Instead of targeting the world at large and marketing an exciting but early piece of technology with slick concept videos, Mojo is opting to go after niche markets—which AR headsets like Google Glass and Magic Leap have also pivoted toward—and waited until they had something solid to announce with clear next steps.

“We’re really confident about this working,” VP of product and marketing Steve Sinclair told Fast Company. “That’s why we’ve come out of stealth, because we’re seeing all the pieces coming together into a product that does everything we want it to do.”

While there are some tech geeks out there that will want these lenses yesterday—and no doubt plenty of digital skeptics who think our screen addiction is quite personal enough already—few of them will be able to try a pair for years to come.

Mojo says their first priority is to help the people who need it most.

Already designated an FDA Breakthrough Device to speed approvals (which, unlike glasses, are necessary for contact lenses), they’ll aim their product at folks with severely impaired vision. The lenses will use edge detection to highlight people and things and a zoom function to enlarge objects. The next step may be special applications in professions where hands free information is critical, like firefighters coordinating a rescue in a burning house.

Only after these steps will Mojo move on to a product the average person might wear.

This cautious approach is merited and likely influenced by recent events in the space. Other smart contact lenses, such as Verily’s glucose sensing lens, haven’t made it out of the research phase. And questions of privacy, design, and safety will have to be answered for the product to be accepted and widely adopted. All of which may take a while.

The Next Great Interface (Is Still in Development)

In 2012, legendary video game developer John Carmack strapped a pair of duct-taped ski goggles onto a line of awed journalists at the E3 gaming conference. Inside the goggles—which had been invented by 19-year-old engineer and VR enthusiast, Palmer Luckey—was a digital world that shifted with your gaze, just like the real world would. It was immersive VR, the best yet for a product of its size and cost, and resulted in the $2 billion acquisition of Oculus by Facebook.

That story is now enshrined in tech lore. A kid, a garage, a prototype that sets the world on fire.

But the original Oculus Rift wasn’t anywhere close to a finished product. And the finished product wasn’t anywhere near a must-have device. It took several years for the first consumer VR headsets to come out and nearly as long since to begin denting the mainstream.

Recently, we’ve seen high-quality VR headsets go wireless, do away with costly gaming laptops, and even morph into steampunk glasses. But the latest generation of VR is still working out its potential almost a decade after its birth.

Likewise, the Mojo Lens isn’t an imminent consumer product, and because the lenses go in your eyes, the hurdles ahead are a lot higher than those faced by makers of VR headsets.

Clinical trials won’t begin for a couple of years, and the company plans to hold off on a mainstream offering until after its work with the FDA and niche industry applications. There are some, too, who wonder whether hard scleral lenses make sense for the general public beyond medical use, both in terms of comfort and care (keeping them clean, especially). If a mainstream version needs flexible components, it might be a long wait indeed.

Still, the Mojo Lens is a real and tangible step forward. A prototype, yes, but a prototype showing that key components, like displays, sensors, radios, and chips, can now be made small and efficient enough to be combined on a contact lens.

It may well be the starting pistol in a race to AR on your eye. And however long the race takes or whatever the future holds for Mojo specifically, that’s a pretty wild thought.

Image Credit: Mojo Vision

Source: Singularity Hub:  https://singularityhub.com/2020/01/17/mojo-visions-augmented-reality-contact-lenses-kick-off-a-race-to-ar-in-your-eye/

The Future of Money: AI Investors, Crowdlending, and the Death of Cash

· · Posted in Artificial Intelligence, Experts, Future, General, Science and Society ·

The Future of Money: AI Investors, Crowdlending, and the Death of Cash

Every day, roughly 60 percent of all market trades are made by computer. When the market turns volatile, this can climb to as high as 90 percent. Robo-advisors are increasingly making this process available to the consumer, saving them time and money as a result.

With humans no longer in the transaction chain, fees are slashed. Undercutting the typical 2 percent cut of profits (not to mention 20 percent incentives) charged by a wealth manager, most robo-advisors take around 0.25 percent. And AI is inserting itself into almost every aspect of the financial world.

“Fintech” describes the convergence of technology and financial services. First colonized by networks and apps, it was then radicalized by AI and blockchain, and now underpins a global wealth redistribution mechanism.

Today, we’ll discuss the ways in which AI algorithms are already massively disrupting financial systems across the globe.

Let’s dive in.

Crowdlending

Wherever large amounts of customer frustration encounter large piles of money, opportunity lurks.

This gave rise to a company called TransferWise. By matching customers who have, say, pesos that they want to turn into dollars with customers who want to change dollars into pesos, TransferWise is using a modified dating app to take on the entire foreign currency exchange market. In fact, the company reached a $3.5 billion valuation in under five years.

Built on networks and apps, TransferWise is also an example of fintech’s colonization wave. The radicalized wave arose when AI entered the picture.

Consider the age-old practice of “Buddy, can I borrow a dollar,” otherwise known as peer-to-peer lending. Traditionally, this is a high-risk practice—which is to say, Buddy rarely gets his dollar back.

This problem only gets worse with scale. As villages turned into towns, towns expanded into cities, and cities began to sprawl, neighborly trust broke down. That’s where banks came into play—they added trust back into the lending equation.

But who needs trust when there’s data?

With AI, huge groups of people can come together, share financial information, and pool risk, becoming the peer-to-peer market now known as “crowdlending.”

Prosper, Funding Circle, and LendingTree are three example players in the market, a market expected to surge from $26.16 billion in 2015 to $897.85 billion by 2024.

A different example is the Smart Finance Group. Created in 2013 to serve China’s massive unbanked and underbanked population, Smart Finance uses an AI to comb a user’s personal data—social media data, smartphone data, educational and employment history, etc.—to generate a reliable credit score nearly instantly.

With this method, they can approve a peer-to-peer loan in under eight seconds, including micro-loans to the unbanked.

And the results speak for themselves. Roughly 1.5 to 2 million loans are taken out every month via Smart Finance.

Investing with AI

AI is also making an impact on investing.

Traditionally, this game was played by the wealthy as it is a game of data. Financial advisors had the best data, but you needed to be wealthy enough to afford a financial advisor to access it.

And advisors are picky. Since it can take more time to manage small investors than large investors, many wealth managers have investment minimums in the range of hundreds of thousands of dollars.

But AI has leveled the playing field.

Today, robo-advisors like Wealthfront and Betterment are bringing wealth management to the masses. Via an app, clients answer a series of initial questions about risk tolerances, investment goals, and retirement aims, and then algorithms take over.

Actually, algorithms have already taken over.

As mentioned earlier, up to 90 percent of market trades are made by computer. All robo-advisors have done is make the process available directly to customers, saving them money in the process. Instead of the typical 2 percent of profits (and 20 percent incentives) charged by a wealth manager, most robo-advisors take around a quarter of a percent.

And investors are responding.

As of January 2019, Wealthfront had $11 billion under management, while Betterment stood at $14 billion. While robo-advisors still account for only roughly 1 percent of total US investment, Business Insider Intelligence estimates that number will climb to $4.6 trillion by 2022.

The Death of Cash

Finally we come to our last category, using money to pay for things.

But we already know this story. When was the last time you dropped coins into a toll booth? Or paid cash for a cab ride? In fact, Uber and Lyft allow us to get around a city without a wallet. Cashier-less stores with services like Amazon Go and Uber Eats and wallet-less pay mechanisms are about to become the new normal.

Denmark stopped printing money in 2017. The year prior, in an attempt to expand mobile banking and demonetize the country’s gray-market economy, India recalled 86 percent of its cash. Vietnam aims to make retail 90 percent cashless by 2020. And Sweden, where over 80 percent of all transactions are already digital, is almost there.

Economists often point out that two of the main factors driving economic growth are the availability of money—the stockpiles we can draw upon—and the velocity of money, or the speed and ease with which we can move that money around. Both of these factors are being amplified by exponential technologies.

As our transactions shift to the digital realm, more data on spending habits can be collected and fed to AI algorithms that continue to learn based on real-time input. This information will educate marketing campaigns, credit records, and investment goals.

Final Thoughts

What, exactly, do we do with our money?

We store it, of course. Mostly in banks. We also move it around, sometimes transferring cash between companies, other times borrowing or lending among individuals. Next, we invest it, attempting to use our money to grow more money. Finally, since the time coins were conch shells, we trade it for the stuff we want.

Thanks to converging exponentials, each of these areas is being reimagined, with bits and bytes replacing dollars and cents. And neither contemporary economics nor the way we live our lives will ever be the same.

Join Me

(1) A360 Executive Mastermind: If you’re an exponentially and abundance-minded entrepreneur who would like coaching directly from me, consider joining my Abundance 360 Mastermind, a highly selective community of 360 CEOs and entrepreneurs who I coach for 3 days every January in Beverly Hills, Ca. Through A360, I provide my members with context and clarity about how converging exponential technologies will transform every industry. I’m committed to running A360 for the course of an ongoing 25-year journey as a “countdown to the Singularity.”

If you’d like to learn more and consider joining our 2020 membership, apply here.

(2) Abundance-Digital Online Community: I’ve also created a Digital/Online community of bold, abundance-minded entrepreneurs called Abundance-Digital. Abundance-Digital is Singularity University’s ‘onramp’ for exponential entrepreneurs — those who want to get involved and play at a higher level. Click here to learn more.

(Both A360 and Abundance-Digital are part of Singularity University — your participation opens you to a global community.)

This article originally appeared on diamandis.com. Read the original article here.

Image Credit: Image by Lorenzo Cafaro from Pixabay

Source: Singularity Hub:  https://singularityhub.com/2020/01/16/the-future-of-money-ai-investors-crowdlending-and-the-death-of-cash/

How Google’s New Weather AI Will Make Sure You Never Get Caught in the Rain

· · Posted in Artificial Intelligence, Creative Commons, General, Science and Society, Topics ·

How Google’s New Weather AI Will Make Sure You Never Get Caught in the Rain

Among the many things we’ve become addicted to on our smartphones is checking the weather. If you’re anything like me, you open a weather app at least twice a day: in the morning to know what to expect for the day ahead, maybe before your commute home so you can prepare for possible rain or snow, and sometimes before bed to get an idea of what to wear or what activities to plan for the next day. Depending where you live, how much time you spend outside, and how prone your area is to rapid weather changes, maybe you check the forecast even more frequently than that.

The fact that our phones now contain hour-by-hour breakdowns of temperature and likelihood of precipitation means we can be well-informed and well-prepared. But these forecasts are coming at a greater cost than we know, and they’re not always right.

On Monday, a post on Google’s AI blog shared a machine learning method the company is developing for weather prediction. Google calls the technique “nowcasting” because it’s set up to predict weather zero to six hours ahead of time, and is focused around weather events like thunderstorms that can quickly morph from clear skies to heavy rains to gusting wind and back again.

By simplifying its methodology and actually using less data than existing forecasting techniques, the company believes it can give us accurate, timely weather predictions, especially ones relating to precipitation.

One of the biggest problems with existing forecasting is the amount of data it incorporates, and the amount of computing power needed to process and make sense of all that data. The US National Oceanic and Atmospheric Administration (NOAA) alone collects 100 terabytes of data per day. It’s input to forecasting models that simulate everything from atmospheric dynamics and thermal radiation to lake and ocean effects.

The forecasting engines run on supercomputers, but they’re limited by their heavy need for computing power. According to Google’s blog post, computational demands limit weather prediction to a spatial resolution of about 5 kilometers (3.1 miles), and the simulations take several hours to run. If, for example, it takes six hours to compute a forecast, that allows only three to four runs per day and results in forecasts based on data that’s six hours old or more.

Google presents its method as being far simpler, describing it as highly localized, data-driven, physics-free, and low-latency. Essentially, the method turns weather forecasting into a computer vision problem; based on progressive images of the formation and movement of clouds over a short time period, a machine learning algorithm predicts how the pattern will evolve over the subsequent few hours.

Specifically, Google uses a convolutional neural network (CNN), a type of deep learning algorithm whose architecture is particularly conducive to image analysis. The “physics-free” descriptor means the neural network learns only from its training data and doesn’t incorporate knowledge of how the atmosphere works; all it has to go off of are patterns it identifies in the images it’s fed.

Google points out that from a computing power perspective, this method is much more economical than existing forecast techniques, especially once the model is already trained. The company claims its algorithm can generate forecasts that have a one-kilometer resolution with a latency of five to ten minutes.

When compared to three common forecasting models—the NOAA’s high resolution rapid refresh numerical forecast, an optical flow algorithm, and a persistence model—Google claims the precision of its method outperformed all of these for prediction timespans of less than five to six hours, emphasizing that low latency makes its predictions “effectively instantaneous.”

So instead of checking the weather on our phones a couple times a day or every few hours, it may not be long before we can access accurate forecasts every minute.

We could also just look out the window—or put our phones down and go for a walk outside.

Image Credit: Image by _Alicja_ from Pixabay

Source: Singularity Hub:  https://singularityhub.com/2020/01/15/how-googles-new-weather-ai-makes-instant-accurate-forecasts/

Scientists Discovered ‘Mini-Computers’ in Human Neurons—and That’s Great News for AI

· · Posted in General, Neuroscience, Science and Society, Topics ·

Scientists Discovered ‘Mini-Computers’ in Human Neurons—and That’s Great News for AI

With just their input cables, human neurons can perform difficult logic calculations previously only seen in entire neural networks. To restate: human neurons are far more powerful devices than originally thought. And if deep learning algorithms—the AI method loosely based on the brain that’s taken our world by storm—take note, they can be too.

Those are unconventional, fighting words.

For 70 years, neurons were considered the basic computational unit of the brain. Yet according to a new study published this month in Science, the neurons in our cortex, the outermost “crust” of our brain, seem to have uniquely evolved to sustain incredibly complex computations in their input cables. It’s as if someone finally obtained proof that your computer’s electrical wiring is actually made up of mini-processors, each performing calculations before sending results to a CPU.

It’s weird. It’s controversial. But it has also just been seen for the first time in human neurons.

As the authors conclude: we long assumed that a neuron could only operate logical functions such as AND and OR, whereas more complex computations required entire networks. We find that activity in a neuron’s input cables can support complex logical operations using completely different rules than a single neuron.

So why should we care? Fundamentally, it has to do with intelligence—why we stand out among the animal kingdom, and how we can potentially replicate that intelligence with AI.

Like the Earth’s crust, the cortex is also made up of multiple layers, with distinctive wiring patterns that link up neurons within layers and among different ones. Neuroscientists have long thought that our enormously intricate cortex contributes to our intellectual capabilities—in fact, deep learning was inspired by computations embedded within cortical neurons.

But the new results, recorded from surgically-removed brain chunks from patients with brain tumors and epilepsy, suggest that current deep learning methods are only scratching the surface of replicating our brain’s computations. If AI systems can incorporate these newly discovered algorithms, they could potentially become far more powerful.

Meet the All-or-None Neuron

A textbook neuron looks like a leafless tree: massive roots, called dendrites, lead to a sturdy, bulbous base—the body. Like water and nutrients, incoming electrical signals shoot up dendritic roots into the body, where a hump-like structure synthesizes all the information. If the stimulation is sufficiently strong, it gets passed down a singular tree trunk—the output cable called an axon—then transmitted to another neuron by way of bubbles filled with chemical messengers or with electricity. If the input signals are too weak, the neuron kills the data. It’s why neuroscientists often call single neurons “binary” or “digital”: they either fire or don’t.

Simple, no?

Well…not quite. For decades, a question nagged at the back of neuroscientists’ minds: why are dendritic trees, compared to a single lonely axon, so much more intricate?

By recording from single neurons in rodent brains, scientists recently began figuring out that dendritic trees aren’t just simple passive cables. Rather, they’re extremely active components underlying a hidden layer of neural computation. Some dendritic trees, for example, can generate electrical spikes five times larger and more frequently than classic neuronal firing. Just in rats, the discovery of active dendrites mean that the brain could have 100 times more processing capacity than previously thought.

The new study asks: does the same hold true for humans?

Human Dendrites Are Special

Compared to rodent brains, the multi-layered human cortex is much thicker and denser. Layers 2 and 3 (L2/3) especially stand out for their elaborate and densely-packed dendritic forests. Compared to other species—or even the rest of the human brain—these layers contain a disproportionate amount of neuronal matter. The root cause of this strange thickening lies in our genes, which encode a brain development program to guide the characteristic. Some even believe that it’s fundamental to what makes us human.

If dendrite “inputs” help shape our neurons’ computation—and our intelligence—then L2/3 is where we should be able to observe them, the authors reasoned.

Measuring electrical activity from dendrites, each 100 times smaller than the diameter of a human hair, is much easier said than done. It’s partly why these enormously powerful calculations have been hard to capture using electrodes even in animals—the process is similar to gently sucking on an ant’s back with a Roman column-sized straw without hurting the ant.

Rather than recording from a living, intact human brain, the team opted to look at fresh slices of the cortex removed due to epilepsy or tumors. It’s a smart strategy: slices are much easier to examine using traditional neuroscience methods—for example, something called a “patch clamp” that records directly from neuronal components. Slices can also be examined under the microscope using fluorescent dyes that glow during activity. Using brain tissue from two different types of patients can then help weed out signals unique to each brain disease to get to the root of human dendritic computations.

A bizarre signal immediately emerged. Human dendrites sparked with activity, but the electrical spikes quickly dissipated as they traveled towards the cell body. In contrast, a standard neural signal doesn’t taper down as it gallops along the output cable towards its next destination. Even weirder, the dendritic signals relied strictly on calcium ions to generate their electricity, which massively differs from classic neural signaling.

It’s like suddenly discovering a new species that consumes carbon dioxide, rather than oxygen, to sustain its activity—except that species is part of you. These signals, dubbed “dCaAPs,” have never been observed in cortical cells from any mammals previously, the authors said.

“There was a ‘eureka’ moment when we saw the dendritic action potentials for the first time,” said study co-author Dr. Matthew Larkum at Humboldt University of Berlin. “The experiments were very challenging, so to push the questions past just repeating what has been done in rodents already was very satisfying.”

But it gets weirder. Unlike a neuron’s all-or-none firing, human dendrites seem to go analogue. That is, their response is “graded,” but in an unintuitive way: the stronger their stimuli, the lower their response. This is in stark contrast to other neuronal computations, where stronger input, even from multiple sources, usually leads to stronger output. And while these dendritic spikes aren’t loners per se—a few dCaAPs helped change the firing of its neuron—many of the dendrite’s electrical activity seemed to do their own thing.

Forest in the Trees

Cataloging the secret lives of human dendrites is already interesting, but the authors went a step further to ask what it all means.

Using computational modeling, they recreated dCaAPs’ unique firing pattern and challenged it to solve a logic function called XOR. It compares two inputs, and if the bits are the same, the result is 0. If they’re different, it results in 1. Unlike the simpler AND and OR functions, XOR normally requires an entire neural network to perform.

However, human dendrites’ strange behavior, where one input only leads to one output, allowed them to “effectively compute the XOR operation,” the authors said. When stacked together with a neuron’s normal AND and OR functions, it’s then possible to condense entire network functions into that of a single neuron. However, for now the idea remains theoretical—the authors weren’t able to model an entire neuron along with dendritic computations.

But keep your eye out for updates. The results, if validated in intact human brains, hold enormous possibilities for improving deep learning algorithms. For now, deep learning uses individual artificial “neurons” that link into multi-layered networks—similar to our previous understanding of human brains. Adding dendritic computations could in theory massively expand deep learning capabilities. In a way, AI is now neuroscience’s theoretical playground, a collaboration made in heaven.

Regardless, the results peel back another onion layer towards understanding and replicating our intelligence. “Dendrites make up 95 percent of the surface area of pyramidal cells in the cortex, but have remained ‘unexplored territory’ in the human brain,” said Dr. Michael Häusser at University College London, who was not involved in the study. By hunting for similar signals in rodent brains, we may be able to determine whether “the special electrical properties of human dendrites play a key role in making human brains special,” he said.

Image Credit: Image by Gerd Altmann from Pixabay

Source: Singularity Hub:  https://singularityhub.com/2020/01/14/scientists-discovered-mini-computers-in-human-neurons-and-thats-great-news-for-ai/

IBM’s Plan to Design Solid-State Batteries Using Quantum Tech

· · Posted in Computing, Energy, General, Science and Society, Topics ·

IBM’s Plan to Design Solid-State Batteries Using Quantum Tech

Batteries are the key to decarbonizing both transport and the grid, but today’s technology is still a long way from living up to this promise. IBM seems to have decided its computing chops are the key to solving the problem.

Lithium-ion batteries are still the gold standard technology in this field, and they’ve come a long way; 10 years ago they could just about get your iPod through the day, today they can power high-performance cars over hundreds of miles.

But if we want to reach a point where batteries can outperform gasoline or store huge amounts of solar energy, we need some breakthroughs. So IBM has teamed up with Mercedes-Benz and its parent company Daimler to develop new batteries that could match up to our needs.

At the annual Consumer Electronics Show (CES) last week, they demonstrated a proof-of-concept showing how IBM’s quantum computers could help design cutting-edge solid-state batteries. At the end of last year they touted a new battery made from materials that can be extracted from the sea, which outperforms today’s technology across the board.

The latter breakthrough was met with some skepticism, because the company chose to announce it via a blog post rather than published research. But while crucial details regarding the exact composition of the battery have been left out, the claimed specs are impressive.

IBM says its battery combines three new materials (which it hasn’t disclosed) that can be optimized for a host of different characteristics, including lower cost, faster charging time, higher power and energy density, energy efficiency, and low flammability.

When configured for high power it can charge to 80 percent in just 5 minutes, which would make charging an electric vehicle as simple as filling a gas tank. It can also match the energy density of state-of-the-art lithium-ion batteries, but outperforms them when it comes to power density, which will be crucial for applications like electric planes that require big bursts of power. The chemistry can also be tweaked for long life cycle, which is crucial for grid storage.

IBM was keen to stress that unlike most existing batteries, this new design contains no heavy metals, like nickel or cobalt, that are extracted in environmentally damaging ways and often under questionable working conditions. The researchers say the materials could even be extracted from sea water, though that would likely require development of new technology.

The researchers told IEEE Spectrum that they have teamed up with with Mercedes Benz R&D North America, Japanese chemical company Central Glass, and battery startup Sidus to start commercially developing the battery, which could come to fruition in one to two years.

And that’s not the only collaboration IBM has going with the German car maker. Scientists at IBM and Daimler have joined forces to uses the computing giant’s quantum computers to probe the properties of three molecules that could form in the operation of lithium-sulfur batteries.

This battery chemistry promises to be more powerful, longer-lasting, and cheaper than lithium ion technology and could also make it possible to create solid-state batteries that are, in theory, lighter and more compact. But it’s still experimental at this stage, and there are plenty of mysteries about how all the components interact.

Simulating the three molecules will help us better understand how their behavior will affect important properties like energy storage and discharge, but it can take huge amounts of computing power. Quantum computers hold the promise of doing these kinds of simulations much more efficiently.

However, this piece of research was more of a proof-of-concept than a breakthrough in battery design, the researchers say in a blog post, as the simulations were run on both quantum and classical computers and compared. The quantum simulations came out favorably, which suggests that as IBM’s quantum processors become more powerful they will be able to make fundamental progress in battery design.

How soon that will be remains uncertain, but if we’re banking on battery breakthroughs to solve the climate crisis, it’s good news that Big Blue is on board.

Image Credit: Image courtesy of IBM Research

Source: Singularity Hub:  https://singularityhub.com/2020/01/13/ibms-plan-to-design-solid-state-batteries-using-quantum-tech/