Tag Archives: never

#437373 Microsoft’s New Deepfake Detector Puts ...

The upcoming US presidential election seems set to be something of a mess—to put it lightly. Covid-19 will likely deter millions from voting in person, and mail-in voting isn’t shaping up to be much more promising. This all comes at a time when political tensions are running higher than they have in decades, issues that shouldn’t be political (like mask-wearing) have become highly politicized, and Americans are dramatically divided along party lines.

So the last thing we need right now is yet another wrench in the spokes of democracy, in the form of disinformation; we all saw how that played out in 2016, and it wasn’t pretty. For the record, disinformation purposely misleads people, while misinformation is simply inaccurate, but without malicious intent. While there’s not a ton tech can do to make people feel safe at crowded polling stations or up the Postal Service’s budget, tech can help with disinformation, and Microsoft is trying to do so.

On Tuesday the company released two new tools designed to combat disinformation, described in a blog post by VP of Customer Security and Trust Tom Burt and Chief Scientific Officer Eric Horvitz.

The first is Microsoft Video Authenticator, which is made to detect deepfakes. In case you’re not familiar with this wicked byproduct of AI progress, “deepfakes” refers to audio or visual files made using artificial intelligence that can manipulate peoples’ voices or likenesses to make it look like they said things they didn’t. Editing a video to string together words and form a sentence someone didn’t say doesn’t count as a deepfake; though there’s manipulation involved, you don’t need a neural network and you’re not generating any original content or footage.

The Authenticator analyzes videos or images and tells users the percentage chance that they’ve been artificially manipulated. For videos, the tool can even analyze individual frames in real time.

Deepfake videos are made by feeding hundreds of hours of video of someone into a neural network, “teaching” the network the minutiae of the person’s voice, pronunciation, mannerisms, gestures, etc. It’s like when you do an imitation of your annoying coworker from accounting, complete with mimicking the way he makes every sentence sound like a question and his eyes widen when he talks about complex spreadsheets. You’ve spent hours—no, months—in his presence and have his personality quirks down pat. An AI algorithm that produces deepfakes needs to learn those same quirks, and more, about whoever the creator’s target is.

Given enough real information and examples, the algorithm can then generate its own fake footage, with deepfake creators using computer graphics and manually tweaking the output to make it as realistic as possible.

The scariest part? To make a deepfake, you don’t need a fancy computer or even a ton of knowledge about software. There are open-source programs people can access for free online, and as far as finding video footage of famous people—well, we’ve got YouTube to thank for how easy that is.

Microsoft’s Video Authenticator can detect the blending boundary of a deepfake and subtle fading or greyscale elements that the human eye may not be able to see.

In the blog post, Burt and Horvitz point out that as time goes by, deepfakes are only going to get better and become harder to detect; after all, they’re generated by neural networks that are continuously learning from and improving themselves.

Microsoft’s counter-tactic is to come in from the opposite angle, that is, being able to confirm beyond doubt that a video, image, or piece of news is real (I mean, can McDonald’s fries cure baldness? Did a seal slap a kayaker in the face with an octopus? Never has it been so imperative that the world know the truth).

A tool built into Microsoft Azure, the company’s cloud computing service, lets content producers add digital hashes and certificates to their content, and a reader (which can be used as a browser extension) checks the certificates and matches the hashes to indicate the content is authentic.

Finally, Microsoft also launched an interactive “Spot the Deepfake” quiz it developed in collaboration with the University of Washington’s Center for an Informed Public, deepfake detection company Sensity, and USA Today. The quiz is intended to help people “learn about synthetic media, develop critical media literacy skills, and gain awareness of the impact of synthetic media on democracy.”

The impact Microsoft’s new tools will have remains to be seen—but hey, we’re glad they’re trying. And they’re not alone; Facebook, Twitter, and YouTube have all taken steps to ban and remove deepfakes from their sites. The AI Foundation’s Reality Defender uses synthetic media detection algorithms to identify fake content. There’s even a coalition of big tech companies teaming up to try to fight election interference.

One thing is for sure: between a global pandemic, widespread protests and riots, mass unemployment, a hobbled economy, and the disinformation that’s remained rife through it all, we’re going to need all the help we can get to make it through not just the election, but the rest of the conga-line-of-catastrophes year that is 2020.

Image Credit: Darius Bashar on Unsplash Continue reading

Posted in Human Robots

#437303 The Deck Is Not Rigged: Poker and the ...

Tuomas Sandholm, a computer scientist at Carnegie Mellon University, is not a poker player—or much of a poker fan, in fact—but he is fascinated by the game for much the same reason as the great game theorist John von Neumann before him. Von Neumann, who died in 1957, viewed poker as the perfect model for human decision making, for finding the balance between skill and chance that accompanies our every choice. He saw poker as the ultimate strategic challenge, combining as it does not just the mathematical elements of a game like chess but the uniquely human, psychological angles that are more difficult to model precisely—a view shared years later by Sandholm in his research with artificial intelligence.

“Poker is the main benchmark and challenge program for games of imperfect information,” Sandholm told me on a warm spring afternoon in 2018, when we met in his offices in Pittsburgh. The game, it turns out, has become the gold standard for developing artificial intelligence.

Tall and thin, with wire-frame glasses and neat brow hair framing a friendly face, Sandholm is behind the creation of three computer programs designed to test their mettle against human poker players: Claudico, Libratus, and most recently, Pluribus. (When we met, Libratus was still a toddler and Pluribus didn’t yet exist.) The goal isn’t to solve poker, as such, but to create algorithms whose decision making prowess in poker’s world of imperfect information and stochastic situations—situations that are randomly determined and unable to be predicted—can then be applied to other stochastic realms, like the military, business, government, cybersecurity, even health care.

While the first program, Claudico, was summarily beaten by human poker players—“one broke-ass robot,” an observer called it—Libratus has triumphed in a series of one-on-one, or heads-up, matches against some of the best online players in the United States.

Libratus relies on three main modules. The first involves a basic blueprint strategy for the whole game, allowing it to reach a much faster equilibrium than its predecessor. It includes an algorithm called the Monte Carlo Counterfactual Regret Minimization, which evaluates all future actions to figure out which one would cause the least amount of regret. Regret, of course, is a human emotion. Regret for a computer simply means realizing that an action that wasn’t chosen would have yielded a better outcome than one that was. “Intuitively, regret represents how much the AI regrets having not chosen that action in the past,” says Sandholm. The higher the regret, the higher the chance of choosing that action next time.

It’s a useful way of thinking—but one that is incredibly difficult for the human mind to implement. We are notoriously bad at anticipating our future emotions. How much will we regret doing something? How much will we regret not doing something else? For us, it’s an emotionally laden calculus, and we typically fail to apply it in quite the right way. For a computer, it’s all about the computation of values. What does it regret not doing the most, the thing that would have yielded the highest possible expected value?

The second module is a sub-game solver that takes into account the mistakes the opponent has made so far and accounts for every hand she could possibly have. And finally, there is a self-improver. This is the area where data and machine learning come into play. It’s dangerous to try to exploit your opponent—it opens you up to the risk that you’ll get exploited right back, especially if you’re a computer program and your opponent is human. So instead of attempting to do that, the self-improver lets the opponent’s actions inform the areas where the program should focus. “That lets the opponent’s actions tell us where [they] think they’ve found holes in our strategy,” Sandholm explained. This allows the algorithm to develop a blueprint strategy to patch those holes.

It’s a very human-like adaptation, if you think about it. I’m not going to try to outmaneuver you head on. Instead, I’m going to see how you’re trying to outmaneuver me and respond accordingly. Sun-Tzu would surely approve. Watch how you’re perceived, not how you perceive yourself—because in the end, you’re playing against those who are doing the perceiving, and their opinion, right or not, is the only one that matters when you craft your strategy. Overnight, the algorithm patches up its overall approach according to the resulting analysis.

There’s one final thing Libratus is able to do: play in situations with unknown probabilities. There’s a concept in game theory known as the trembling hand: There are branches of the game tree that, under an optimal strategy, one should theoretically never get to; but with some probability, your all-too-human opponent’s hand trembles, they take a wrong action, and you’re suddenly in a totally unmapped part of the game. Before, that would spell disaster for the computer: An unmapped part of the tree means the program no longer knows how to respond. Now, there’s a contingency plan.

Of course, no algorithm is perfect. When Libratus is playing poker, it’s essentially working in a zero-sum environment. It wins, the opponent loses. The opponent wins, it loses. But while some real-life interactions really are zero-sum—cyber warfare comes to mind—many others are not nearly as straightforward: My win does not necessarily mean your loss. The pie is not fixed, and our interactions may be more positive-sum than not.

What’s more, real-life applications have to contend with something that a poker algorithm does not: the weights that are assigned to different elements of a decision. In poker, this is a simple value-maximizing process. But what is value in the human realm? Sandholm had to contend with this before, when he helped craft the world’s first kidney exchange. Do you want to be more efficient, giving the maximum number of kidneys as quickly as possible—or more fair, which may come at a cost to efficiency? Do you want as many lives as possible saved—or do some take priority at the cost of reaching more? Is there a preference for the length of the wait until a transplant? Do kids get preference? And on and on. It’s essential, Sandholm says, to separate means and the ends. To figure out the ends, a human has to decide what the goal is.

“The world will ultimately become a lot safer with the help of algorithms like Libratus,” Sandholm told me. I wasn’t sure what he meant. The last thing that most people would do is call poker, with its competition, its winners and losers, its quest to gain the maximum edge over your opponent, a haven of safety.

“Logic is good, and the AI is much better at strategic reasoning than humans can ever be,” he explained. “It’s taking out irrationality, emotionality. And it’s fairer. If you have an AI on your side, it can lift non-experts to the level of experts. Naïve negotiators will suddenly have a better weapon. We can start to close off the digital divide.”

It was an optimistic note to end on—a zero-sum, competitive game yielding a more ultimately fair and rational world.

I wanted to learn more, to see if it was really possible that mathematics and algorithms could ultimately be the future of more human, more psychological interactions. And so, later that day, I accompanied Nick Nystrom, the chief scientist of the Pittsburgh Supercomputing Center—the place that runs all of Sandholm’s poker-AI programs—to the actual processing center that make undertakings like Libratus possible.

A half-hour drive found us in a parking lot by a large glass building. I’d expected something more futuristic, not the same square, corporate glass squares I’ve seen countless times before. The inside, however, was more promising. First the security checkpoint. Then the ride in the elevator — down, not up, to roughly three stories below ground, where we found ourselves in a maze of corridors with card readers at every juncture to make sure you don’t slip through undetected. A red-lit panel formed the final barrier, leading to a small sliver of space between two sets of doors. I could hear a loud hum coming from the far side.

“Let me tell you what you’re going to see before we walk in,” Nystrom told me. “Once we get inside, it will be too loud to hear.”

I was about to witness the heart of the supercomputing center: 27 large containers, in neat rows, each housing multiple processors with speeds and abilities too great for my mind to wrap around. Inside, the temperature is by turns arctic and tropic, so-called “cold” rows alternating with “hot”—fans operate around the clock to cool the processors as they churn through millions of giga, mega, tera, peta and other ever-increasing scales of data bytes. In the cool rows, robotic-looking lights blink green and blue in orderly progression. In the hot rows, a jumble of multicolored wires crisscrosses in tangled skeins.

In the corners stood machines that had outlived their heyday. There was Sherlock, an old Cray model, that warmed my heart. There was a sad nameless computer, whose anonymity was partially compensated for by the Warhol soup cans adorning its cage (an homage to Warhol’s Pittsburghian origins).

And where does Libratus live, I asked? Which of these computers is Bridges, the computer that runs the AI Sandholm and I had been discussing?

Bridges, it turned out, isn’t a single computer. It’s a system with processing power beyond comprehension. It takes over two and a half petabytes to run Libratus. A single petabyte is a million gigabytes: You could watch over 13 years of HD video, store 10 billion photos, catalog the contents of the entire Library of Congress word for word. That’s a whole lot of computing power. And that’s only to succeed at heads-up poker, in limited circumstances.

Yet despite the breathtaking computing power at its disposal, Libratus is still severely limited. Yes, it beat its opponents where Claudico failed. But the poker professionals weren’t allowed to use many of the tools of their trade, including the opponent analysis software that they depend on in actual online games. And humans tire. Libratus can churn for a two-week marathon, where the human mind falters.

But there’s still much it can’t do: play more opponents, play live, or win every time. There’s more humanity in poker than Libratus has yet conquered. “There’s this belief that it’s all about statistics and correlations. And we actually don’t believe that,” Nystrom explained as we left Bridges behind. “Once in a while correlations are good, but in general, they can also be really misleading.”

Two years later, the Sandholm lab will produce Pluribus. Pluribus will be able to play against five players—and will run on a single computer. Much of the human edge will have evaporated in a short, very short time. The algorithms have improved, as have the computers. AI, it seems, has gained by leaps and bounds.

So does that mean that, ultimately, the algorithmic can indeed beat out the human, that computation can untangle the web of human interaction by discerning “the little tactics of deception, of asking yourself what is the other man going to think I mean to do,” as von Neumann put it?

Long before I’d spoken to Sandholm, I’d met Kevin Slavin, a polymath of sorts whose past careers have including founding a game design company and an interactive art space and launching the Playful Systems group at MIT’s Media Lab. Slavin has a decidedly different view from the creators of Pluribus. “On the one hand, [von Neumann] was a genius,” Kevin Slavin reflects. “But the presumptuousness of it.”

Slavin is firmly on the side of the gambler, who recognizes uncertainty for what it is and thus is able to take calculated risks when necessary, all the while tampering confidence at the outcome. The most you can do is put yourself in the path of luck—but to think you can guess with certainty the actual outcome is a presumptuousness the true poker player foregoes. For Slavin, the wonder of computers is “That they can generate this fabulous, complex randomness.” His opinion of the algorithmic assaults on chance? “This is their moment,” he said. “But it’s the exact opposite of what’s really beautiful about a computer, which is that it can do something that’s actually unpredictable. That, to me, is the magic.”

Will they actually succeed in making the unpredictable predictable, though? That’s what I want to know. Because everything I’ve seen tells me that absolute success is impossible. The deck is not rigged.

“It’s an unbelievable amount of work to get there. What do you get at the end? Let’s say they’re successful. Then we live in a world where there’s no God, agency, or luck,” Slavin responded.

“I don’t want to live there,’’ he added “I just don’t want to live there.”

Luckily, it seems that for now, he won’t have to. There are more things in life than are yet written in the algorithms. We have no reliable lie detection software—whether in the face, the skin, or the brain. In a recent test of bluffing in poker, computer face recognition failed miserably. We can get at discomfort, but we can’t get at the reasons for that discomfort: lying, fatigue, stress—they all look much the same. And humans, of course, can also mimic stress where none exists, complicating the picture even further.

Pluribus may turn out to be powerful, but von Neumann’s challenge still stands: The true nature of games, the most human of the human, remains to be conquered.

This article was originally published on Undark. Read the original article.

Image Credit: José Pablo Iglesias / Unsplash Continue reading

Posted in Human Robots

#437282 This Week’s Awesome Tech Stories From ...

ARTIFICIAL INTELLIGENCE
OpenAI’s Latest Breakthrough Is Astonishingly Powerful, But Still Fighting Its Flaws
James Vincent | The Verge
“What makes GPT-3 amazing, they say, is not that it can tell you that the capital of Paraguay is Asunción (it is) or that 466 times 23.5 is 10,987 (it’s not), but that it’s capable of answering both questions and many more beside simply because it was trained on more data for longer than other programs. If there’s one thing we know that the world is creating more and more of, it’s data and computing power, which means GPT-3’s descendants are only going to get more clever.”

TECHNOLOGY
I Tried to Live Without the Tech Giants. It Was Impossible.
Kashmir Hill | The New York Times
“Critics of the big tech companies are often told, ‘If you don’t like the company, don’t use its products.’ My takeaway from the experiment was that it’s not possible to do that. It’s not just the products and services branded with the big tech giant’s name. It’s that these companies control a thicket of more obscure products and services that are hard to untangle from tools we rely on for everything we do, from work to getting from point A to point B.”

ROBOTICS
Meet the Engineer Who Let a Robot Barber Shave Him With a Straight Razor
Luke Dormehl | Digital Trends
“No, it’s not some kind of lockdown-induced barber startup or a Jackass-style stunt. Instead, Whitney, assistant professor of mechanical and industrial engineering at Northeastern University School of Engineering, was interested in straight-razor shaving as a microcosm for some of the big challenges that robots have faced in the past (such as their jerky, robotic movement) and how they can now be solved.”

LONGEVITY
Can Trees Live Forever? New Kindling in an Immortal Debate
Cara Giaimo | The New York Times
“Even if a scientist dedicated her whole career to very old trees, she would be able to follow her research subjects for only a small percentage of their lives. And a long enough multigenerational study might see its own methods go obsolete. For these reasons, Dr. Munné-Bosch thinks we will never prove’ whether long-lived trees experience senescence…”

BIOTECH
There’s No Such Thing as Family Secrets in the Age of 23andMe
Caitlin Harrington | Wired
“…technology has a way of creating new consequences for old decisions. Today, some 30 million people have taken consumer DNA tests, a threshold experts have called a tipping point. People conceived through donor insemination are matching with half-siblings, tracking down their donors, forming networks and advocacy organizations.”

ETHICS
The Problems AI Has Today Go Back Centuries
Karen Hao | MIT Techology Review
“In 2018, just as the AI field was beginning to reckon with problems like algorithmic discrimination, [Shakir Mohamed, a South African AI researcher at DeepMind], penned a blog post with his initial thoughts. In it he called on researchers to ‘decolonise artificial intelligence’—to reorient the field’s work away from Western hubs like Silicon Valley and engage new voices, cultures, and ideas for guiding the technology’s development.”

INTERNET
AI-Generated Text Is the Scariest Deepfake of All
Renee DiResta | Wired
“In the future, deepfake videos and audiofakes may well be used to create distinct, sensational moments that commandeer a press cycle, or to distract from some other, more organic scandal. But undetectable textfakes—masked as regular chatter on Twitter, Facebook, Reddit, and the like—have the potential to be far more subtle, far more prevalent, and far more sinister.”

Image credit: Adrien Olichon / Unsplash Continue reading

Posted in Human Robots

#437269 DeepMind’s Newest AI Programs Itself ...

When Deep Blue defeated world chess champion Garry Kasparov in 1997, it may have seemed artificial intelligence had finally arrived. A computer had just taken down one of the top chess players of all time. But it wasn’t to be.

Though Deep Blue was meticulously programmed top-to-bottom to play chess, the approach was too labor-intensive, too dependent on clear rules and bounded possibilities to succeed at more complex games, let alone in the real world. The next revolution would take a decade and a half, when vastly more computing power and data revived machine learning, an old idea in artificial intelligence just waiting for the world to catch up.

Today, machine learning dominates, mostly by way of a family of algorithms called deep learning, while symbolic AI, the dominant approach in Deep Blue’s day, has faded into the background.

Key to deep learning’s success is the fact the algorithms basically write themselves. Given some high-level programming and a dataset, they learn from experience. No engineer anticipates every possibility in code. The algorithms just figure it.

Now, Alphabet’s DeepMind is taking this automation further by developing deep learning algorithms that can handle programming tasks which have been, to date, the sole domain of the world’s top computer scientists (and take them years to write).

In a paper recently published on the pre-print server arXiv, a database for research papers that haven’t been peer reviewed yet, the DeepMind team described a new deep reinforcement learning algorithm that was able to discover its own value function—a critical programming rule in deep reinforcement learning—from scratch.

Surprisingly, the algorithm was also effective beyond the simple environments it trained in, going on to play Atari games—a different, more complicated task—at a level that was, at times, competitive with human-designed algorithms and achieving superhuman levels of play in 14 games.

DeepMind says the approach could accelerate the development of reinforcement learning algorithms and even lead to a shift in focus, where instead of spending years writing the algorithms themselves, researchers work to perfect the environments in which they train.

Pavlov’s Digital Dog
First, a little background.

Three main deep learning approaches are supervised, unsupervised, and reinforcement learning.

The first two consume huge amounts of data (like images or articles), look for patterns in the data, and use those patterns to inform actions (like identifying an image of a cat). To us, this is a pretty alien way to learn about the world. Not only would it be mind-numbingly dull to review millions of cat images, it’d take us years or more to do what these programs do in hours or days. And of course, we can learn what a cat looks like from just a few examples. So why bother?

While supervised and unsupervised deep learning emphasize the machine in machine learning, reinforcement learning is a bit more biological. It actually is the way we learn. Confronted with several possible actions, we predict which will be most rewarding based on experience—weighing the pleasure of eating a chocolate chip cookie against avoiding a cavity and trip to the dentist.

In deep reinforcement learning, algorithms go through a similar process as they take action. In the Atari game Breakout, for instance, a player guides a paddle to bounce a ball at a ceiling of bricks, trying to break as many as possible. When playing Breakout, should an algorithm move the paddle left or right? To decide, it runs a projection—this is the value function—of which direction will maximize the total points, or rewards, it can earn.

Move by move, game by game, an algorithm combines experience and value function to learn which actions bring greater rewards and improves its play, until eventually, it becomes an uncanny Breakout player.

Learning to Learn (Very Meta)
So, a key to deep reinforcement learning is developing a good value function. And that’s difficult. According to the DeepMind team, it takes years of manual research to write the rules guiding algorithmic actions—which is why automating the process is so alluring. Their new Learned Policy Gradient (LPG) algorithm makes solid progress in that direction.

LPG trained in a number of toy environments. Most of these were “gridworlds”—literally two-dimensional grids with objects in some squares. The AI moves square to square and earns points or punishments as it encounters objects. The grids vary in size, and the distribution of objects is either set or random. The training environments offer opportunities to learn fundamental lessons for reinforcement learning algorithms.

Only in LPG’s case, it had no value function to guide that learning.

Instead, LPG has what DeepMind calls a “meta-learner.” You might think of this as an algorithm within an algorithm that, by interacting with its environment, discovers both “what to predict,” thereby forming its version of a value function, and “how to learn from it,” applying its newly discovered value function to each decision it makes in the future.

Prior work in the area has had some success, but according to DeepMind, LPG is the first algorithm to discover reinforcement learning rules from scratch and to generalize beyond training. The latter was particularly surprising because Atari games are so different from the simple worlds LPG trained in—that is, it had never seen anything like an Atari game.

Time to Hand Over the Reins? Not Just Yet
LPG is still behind advanced human-designed algorithms, the researchers said. But it outperformed a human-designed benchmark in training and even some Atari games, which suggests it isn’t strictly worse, just that it specializes in some environments.

This is where there’s room for improvement and more research.

The more environments LPG saw, the more it could successfully generalize. Intriguingly, the researchers speculate that with enough well-designed training environments, the approach might yield a general-purpose reinforcement learning algorithm.

At the least, though, they say further automation of algorithm discovery—that is, algorithms learning to learn—will accelerate the field. In the near term, it can help researchers more quickly develop hand-designed algorithms. Further out, as self-discovered algorithms like LPG improve, engineers may shift from manually developing the algorithms themselves to building the environments where they learn.

Deep learning long ago left Deep Blue in the dust at games. Perhaps algorithms learning to learn will be a winning strategy in the real world too.

Image credit: Mike Szczepanski / Unsplash Continue reading

Posted in Human Robots

#437267 This Week’s Awesome Tech Stories From ...

ARTIFICIAL INTELLIGENCE
OpenAI’s New Language Generator GPT-3 Is Shockingly Good—and Completely Mindless
Will Douglas Heaven | MIT Technology Review
“‘Playing with GPT-3 feels like seeing the future,’ Arram Sabeti, a San Francisco–based developer and artist, tweeted last week. That pretty much sums up the response on social media in the last few days to OpenAI’s latest language-generating AI.”

ROBOTICS
The Star of This $70 Million Sci-Fi Film Is a Robot
Sarah Bahr | The New York Times
“Erica was created by Hiroshi Ishiguro, a roboticist at Osaka University in Japan, to be ‘the most beautiful woman in the world’—he modeled her after images of Miss Universe pageant finalists—and the most humanlike robot in existence. But she’s more than just a pretty face: Though ‘b’ is still in preproduction, when she makes her debut, producers believe it will be the first time a film has relied on a fully autonomous artificially intelligent actor.”

VIRTUAL REALITY
My Glitchy, Glorious Day at a Conference for Virtual Beings
Emma Grey Ellis | Wired
“Spectators spent much of the time debating who was real and who was fake. …[Lars Buttler’s] eyes seemed awake and alive in a way that the faces of the other participants in the Zoom call—venture capitalist, a tech founder, and an activist, all of them puppeted by artificial intelligence—were not. ‘Pretty sure Lars is human,’ a (real-person) spectator typed in the in-meeting chat room. ‘I’m starting to think Lars is AI,’ wrote another.”

FUTURE OF FOOD
KFC Is Working With a Russian 3D Bioprinting Firm to Try to Make Lab-Produced Chicken Nuggets
Kim Lyons | The Verge
“The chicken restaurant chain will work with Russian company 3D Bioprinting Solutions to develop bioprinting technology that will ‘print’ chicken meat, using chicken cells and plant material. KFC plans to provide the bioprinting firm with ingredients like breading and spices ‘to achieve the signature KFC taste’ and will seek to replicate the taste and texture of genuine chicken.”

BIOTECH
A CRISPR Cow Is Born. It’s Definitely a Boy
Megan Molteni | Wired
“After nearly five years of research, at least half a million dollars, dozens of failed pregnancies, and countless scientific setbacks, Van Eenennaam’s pioneering attempt to create a line of Crispr’d cattle tailored to the needs of the beef industry all came down to this one calf. Who, as luck seemed sure to have it, was about to enter the world in the middle of a global pandemic.”

GOVERNANCE
Is the Pandemic Finally the Moment for a Universal Basic Income?
Brooks Rainwater and Clay Dillow | Fast Company
“Since February, governments around the globe—including in the US—have intervened in their citizens’ individual financial lives, distributing direct cash payments to backstop workers sidelined by the COVID-19 pandemic. Some are considering keeping such direct assistance in place indefinitely, or at least until the economic shocks subside.”

SCIENCE
How Gödel’s Proof Works
Natalie Wolchover | Wired
“In 1931, the Austrian logician Kurt Gödel pulled off arguably one of the most stunning intellectual achievements in history. Mathematicians of the era sought a solid foundation for mathematics: a set of basic mathematical facts, or axioms, that was both consistent—never leading to contradictions—and complete, serving as the building blocks of all mathematical truths. But Gödel’s shocking incompleteness theorems, published when he was just 25, crushed that dream.”

Image credit: Pierre Châtel-Innocenti / Unsplash Continue reading

Posted in Human Robots