AI must read… waitbutwhy?

And here’s where we get to an intense concept: recursive self-improvement. It works like this—

An AI system at a certain level—let’s say human village idiot—is programmed with the goal of improving its own intelligence. Once it does, it’s smarter—maybe at this point it’s at Einstein’s level—so now when it works to improve its intelligence, with an Einstein-level intellect, it has an easier time and it can make bigger leaps. These leaps make it much smarter than any human, allowing it to make even bigger leaps. As the leaps grow larger and happen more rapidly, the AGI soars upwards in intelligence and soon reaches the superintelligent level of an ASI system. This is called an Intelligence Explosion,11 and it’s the ultimate example of The Law of Accelerating Returns.

There is some debate about how soon AI will reach human-level general intelligence. The median year on a survey of hundreds of scientists about when they believed we’d be more likely than not to have reached AGI was 204012—that’s only 25 years from now, which doesn’t sound that huge until you consider that many of the thinkers in this field think it’s likely that the progression from AGI to ASI happens very quickly. Like—this could happen:

It takes decades for the first AI system to reach low-level general intelligence, but it finally happens. A computer is able to understand the world around it as well as a human four-year-old. Suddenly, within an hour of hitting that milestone, the system pumps out the grand theory of physics that unifies general relativity and quantum mechanics, something no human has been able to definitively do. 90 minutes after that, the AI has become an ASI, 170,000 times more intelligent than a human.

Superintelligence of that magnitude is not something we can remotely grasp, any more than a bumblebee can wrap its head around Keynesian Economics. In our world, smart means a 130 IQ and stupid means an 85 IQ—we don’t have a word for an IQ of 12,952.

What we do know is that humans’ utter dominance on this Earth suggests a clear rule: with intelligence comes power. Which means an ASI, when we create it, will be the most powerful being in the history of life on Earth, and all living things, including humans, will be entirely at its whim—and this might happen in the next few decades.

If our meager brains were able to invent wifi, then something 100 or 1,000 or 1 billion times smarter than we are should have no problem controlling the positioning of each and every atom in the world in any way it likes, at any time—everything we consider magic, every power we imagine a supreme God to have will be as mundane an activity for the ASI as flipping on a light switch is for us. Creating the technology to reverse human aging, curing disease and hunger and even mortality, reprogramming the weather to protect the future of life on Earth—all suddenly possible. Also possible is the immediate end of all life on Earth. As far as we’re concerned, if an ASI comes to being, there is now an omnipotent God on Earth—and the all-important question for us is: Will it be a nice God?

https://waitbutwhy.com/2015/01/artificial-intelligence-revolution-1.html

 

Seeding organs… watch this space…

this is one of the most exciting stories of the past few days. Although it is not close to clinical applications yet, it represents a tremendous research direction. The most fascinating element of it to me is how the body knows how to build a liver. We just need to learn to harness the wisdom that is already in the body…

To grow a liver, researchers led by MIT engineer Sangeeta Bhatia started by carefully designing a cellular scaffold for the organ to grow on. They first got human liver cells (hepatocytes) and connective tissue cells (fibroblasts) to grow together in clumps. Then they used a micro-tissue molding to create ropes endothelial cells, which make up the lining of blood and lymphatic vessels. Last, they carefully assembled rows of the cell clumps in between strands of endothelial chords and held the structure together with a biodegradable hydrogel.

In all, they called the organ starter kit SEED, for “In Situ Expansion of Engineered Devices.”

To test out the SEEDs, the researchers implanted them into the belly fat of healthy mice and mice with a genetic disorder that causes liver damage. In the healthy mice, the liver seeds didn’t grow very much. But in the rodents with liver damage—which were circulating liver-regenerating growth factors and other molecular signals to repair their damaged liver—the organ SEEDs sprouted.

Eighty days after implantation, there was a 50-fold cellular expansion along the SEED’s scaffold. The liver organoid formed precursor bile ducts and contained clusters of red blood cells, suggesting vasculature formation. The organoid also pumped out standard human liver proteins, including albumin and transferrin.

There’s a lot more work to go before researchers have a human-sized, functional liver, but the team is optimistic. “We believe that this work sets the stage for using SEEDs as an alternative strategy for scale-up of engineered organs, one that uses native developmental, injury, or regenerative signals to expand prefabricated constructs in situ,” they conclude.

https://arstechnica.com/science/2017/07/mice-grow-their-own-miniature-human-livers-with-organ-starter-kit/

Position Update…

So as of today at 9am for bond hedge /zn and 1pm for gold, my macro hedges are back on, seeking to buffer my positions from Knightian uncertainty. These positions are more a hedge against uncertainty than risk. They aren’t the same thing and I haven’t done a good job of explaining the difference in my portfolio construction. 

In any case, the summer scare is either over or setting us up for a real summer of hurt. I hope to be secure in my process either way… 

Nothing happening here… just quantum teleportation…

Teleportation is a building block for a wide range of technologies. “Long-distance teleportation has been recognized as a fundamental element in protocols such as large-scale quantum networks and distributed quantum computation,” says the Chinese team.

In theory, there should be no maximum distance over which this can be done. But entanglement is a fragile thing because photons interact with matter in the atmosphere or inside optical fibers, causing the entanglement to be lost.

As a result, the distance over which scientists have measured entanglement or performed teleportation is severely limited. “Previous teleportation experiments between distant locations were limited to a distance on the order of 100 kilometers, due to photon loss in optical fibers or terrestrial free-space channels,” says the team.

But Micius changes all that because it orbits at an altitude of 500 kilometers, and for most of this distance, any photons making the journey travel through a vacuum. To minimize the amount of atmosphere in the way, the Chinese team set up its ground station in Ngari in Tibet at an altitude of over 4,000 meters. So the distance from the ground to the satellite varies from 1,400 kilometers when it is near the horizon to 500 kilometers when it is overhead.

To perform the experiment, the Chinese team created entangled pairs of photons on the ground at a rate of about 4,000 per second. They then beamed one of these photons to the satellite, which passed overhead every day at midnight. They kept the other photon on the ground.

Finally, they measured the photons on the ground and in orbit to confirm  that entanglement was taking place, and that they were able to teleport photons in this way. Over 32 days, they sent millions of photons and found positive results in 911 cases. “We report the first quantum teleportation of independent single-photon qubits from a ground observatory to a low Earth orbit satellite—through an up-link channel— with a distance up to 1400 km,” says the Chinese team.

This is the first time that any object has been teleported from Earth to orbit, and it smashes the record for the longest distance for entanglement.

That’s impressive work that sets the stage for much more ambitious goals in the future. “This work establishes the first ground-to-satellite up-link for faithful and ultra-long-distance quantum teleportation, an essential step toward global-scale quantum internet,” says the team.

It also shows China’s obvious dominance and lead in a field that, until recently, was led by Europe and the U.S.—Micius would surely have been impressed. But an important question now is how the West will respond.

Ref: arxiv.org/abs/1707.00934: Ground-to-satellite quantum teleportation

Breakthrough tool predicts properties of theoretical materials

as I have mentioned before, new tools excite me. Because it is out of new tools that our range of possibilities broaden…

Scientists at the University of North Carolina at Chapel Hill and Duke University have created the first general-purpose method for using machine learning to predict the properties of new metals, ceramics and other crystalline materials and to find new uses for existing materials, a discovery that could save countless hours wasted in the trial-and-error process of creating new and better materials.

Researchers led by Olexandr Isayev, Ph.D., and Alexander Tropsha, Ph.D., at the UNC Eshelman School of Pharmacy used data on approximately 60,000 unique materials from the National Institute of Standards and Technology’s Inorganic Crystal Structure Database to create a new methodology they call Properties Labeled Materials Fragments.
Using machine learning to analyze and model existing crystal structures, the PLMF method is able to predict the properties of new materials proposed by scientists and engineers. The tool was even able to fill in missing values for properties of materials in the NIST database that had never been tested experimentally.
“Technology is often driven by the discovery of new materials, but the process of discovering these materials has always been rather haphazard,” Tropsha said. “Out new tool applies the data- and knowledge-driven approach we use in the pharmaceutical sciences to design drugs. Because creating new materials takes an incredible amount of time and effort that often ends in disappointment, our PLMF tool allows materials scientists to test a new idea before they even lift a finger to synthesize it.”

https://phys.org/news/2017-07-breakthrough-tool-properties-theoretical-materials.html

The PLMF method works by creating “fingerprints” from the structure of the crystals that comprise the smallest units of inorganic materials like ceramics, metals and metal alloys. Combining the fingerprints with machine learning allowed the creation of universal models capable of accurately predicting eight critical electronic and thermomechanical properties of virtually any inorganic crystalline material. The properties include conductivity, stiffness and compressibility, heat transfer and response to temperature change, and the team plans to incorporate more properties as they collect more data, Isayev said.
“In many practical projects, people know the range of values they want for a particular property,” Isayev said. “We can leverage what we know about these materials and savvy machine learning to rapidly screen potential materials for the right property. Researchers can quickly narrow candidate materials and avoid many extraneous and complex calculations. This saves money, time and computational resources.”
In the first practical application for the machine learning, the team worked with Assistant Professor Jim Cahoon, Ph.D., in the UNC Department of Chemistry to design a new electrode material for a type of low-cost solar cells. The currently used nickel oxide, is not very efficient, toxic and requires organic solvents to work in the cell.
Scientists virtually screened 50,000 known inorganic compounds and identified lead titanate as the most promising material and subsequent testing confirmed it. The devices using lead titanate exhibited the best performance in aqueous solution, allowing a switch away from solvents to a water-based solution that could help drive down costs while being more environmentally friendly.
“Lead titanate likely would not have been the first choice of most materials scientists because its structure is so dissimilar to nickel oxide,” Isayev said. “Materials derived from iron, cobalt or copper would be more likely to be considered because they are more chemically similar to nickel. The PLMF and machine learning found a simple and novel solutions that saved untold hours of trial-and-error searching.”

 

Read more at: https://phys.org/news/2017-07-breakthrough-tool-properties-theoretical-materials.html#jCp

DARPA Wants Brain Implants That Record From 1 Million Neurons

 

DARPA is known for issuing big challenges. Still, the mission statement for its new Neural Engineering Systems Design program is a doozy: Make neural implants that can record high-fidelity signals from 1 million neurons.

Today’s best brain implants, like the experimental system that a paralyzed man used to control a robotic arm, record from just a few hundred neurons. Recording from 1 million neurons would provide a much richer signal that could be used to better control external devices such as wheelchairs, robots, and computer cursors.

What’s more, the DARPA program calls for the tech to be bidirectional; the implants must be able to not only record signals, but also to transmit computer-generated signals to the neurons. That feature would allow for neural prosthetics that provide blind people with visual information or deaf people with auditory info.

Today the agency announced the six research groups that have been awarded grants under the NESD program. In a press release, DARPA says that even the 1-million-neuron goal is just a starting point. “A million neurons represents a miniscule percentage of the 86 billion neurons in the human brain. Its deeper complexities are going to remain a mystery for some time to come,” says Phillip Alvelda, who launched the program in January. “But if we’re successful in delivering rich sensory signals directly to the brain, NESD will lay a broad foundation for new neurological therapies.”

One of the teams taking on the challenge is the Silicon Valley startup Paradromics. Company CEO Matt Angle says his company is developing a device called the Neural Input-Output Bus (NIOB) that will use bundles of microwire electrodes to interface with neurons. With four bundles containing a total of 200,000 microwires, he says, the NIOB could record from or stimulate 1 million neurons.

“Microwire electrodes have been used since the 1950s, but traditionally they’re un-scaleable,” Angle tells IEEE Spectrum in an interview. With existing systems “you need to wire up one microwire to one amplifier—so if you want to use 100,000 microwires, that’s a lot of soldering work for a grad student,” he says.

Paradromics gets around this problem by polishing the end of a microwire bundle to make it very flat, and then bonding the whole bundle to a chip containing an array of CMOS amplifiers. “We make sure the probability of a single wire coming down and touching the pad on the CMOS is very, very high,” says Angle, “but if you have a few spots that don’t get wires, that doesn’t matter much.”

Each microwire in the bundle has a diameter of less than 20 micrometers.
As always, DARPA emphasizes the practical application of technology. By the end of the four-year NESD program, the teams are expected to have working prototypes that can be used in therapies for sensory restoration.

Paradromics’ goal is a speech prosthetic. The NIOB device’s microwires will record signals from the superior temporal gyrus, a brain area involved in audio processing that decodes speech at the level of sound units called phonemes (other areas of the brain deal with higher-level semantics).

The company drew inspiration from neuroscientist Robert Knight at University of California Berkeley, who has shown that when people read aloud or read silently to themselves the neural signal in the superior temporal gyrus can be used to reconstruct the words. This finding suggests that a user could just imagine speaking a phrase, and a neural implant could record the signal and send the information to a speech synthesizer.

While Paradromics has chosen this speech prosthetic as its DARPA-funded goal, its hardware could be used for any number of neural applications. The differences would come from changing the location of the implant and from the software that decodes the signal.

The challenges ahead of Paradromics are significant. Angle imagines a series of implanted chips, each bonded to 50,000 microwires, that send their data to one central transmitter that sits on the surface of the skull, beneath the skin of the scalp. To deal efficiently with all that data, the implanted system will have to do some processing: “You need to make some decisions inside the body about what you want to send out,” Angle says, “because you can’t have it digitizing and transmitting 50 GB per second.” The central transmitter must then wirelessly send data to a receiver patch worn on the scalp, and must also wirelessly receive power from it.

The other five teams that won NESD grants are research groups investigating vision, speech, and the sense of touch. The group from Brown University, led by neural engineer Arto Nurmikko, is working on a speech prosthetic using tens of thousands of independent “neurograins,” each about the size of a grain of table salt. Those grains will interface with individual neurons, and send their data to one electronics patch that will either be worn on the scalp or implanted under the skin.

In an email, Nurmikko writes that his team is working on such challenges as how to implant the neurograins, how to ensure that they’re hermetically sealed and safe, and how to handle the vast amount of data that they’ll generate. And the biggest challenge of all may be networking 10,000 or 100,000 neurograins together to make one coherent telecommunications system that provides meaningful data.

“Even with a hundred thousand such grains, we would still not reach every neuron—and that’s not the point,” Nurmikko writes. “You want to listen to a sufficiently large number of neurons to understand how, say, the auditory cortex computes ‘the Star Spangled Banner’ for us to have a clear perception of both the music and the words.”

http://spectrum.ieee.org/the-human-os/biomedical/devices/darpa-wants-brain-implants-that-record-from-1-million-neurons

the crazy pace of crispr…

Rewriting the code of life has never been so easy. In 2012 scientists demonstrated a new DNA-editing technique called Crispr. Five years later it is being used to cure mice with HIV and hemophilia. Geneticists are engineering pigs to make them suitable as human organ donors. Bill Gates is spending $75 million to endow a few Anopheles mosquitoes, which spread malaria, with a sort of genetic time bomb that could wipe out the species. A team at Harvard plans to edit 1.5 million letters of elephant DNA to resurrect the woolly mammoth.

“I frankly have been flabbergasted at the pace of the field,” says Jennifer Doudna, a Crispr pioneer who runs a lab at the University of California, Berkeley. “We’re barely five years out, and it’s already in early clinical trials for cancer. It’s unbelievable.”

The thing to understand about Crispr isn’t its acronym—for the record, it stands for Clustered Regularly Interspaced Short Palindromic Repeats—but that it makes editing DNA easy, cheap and precise. Scientists have fiddled with genes for decades, but in clumsy ways. They zapped plants with radiation to flip letters of DNA at random, then looked for useful mutations. They hijacked the infection mechanisms of viruses and bacteria to deliver beneficial payloads. They shot cells with “gene guns,” which are pretty much what they sound like. The first one, invented in the 1980s, was an air pistol modified to fire particles coated with genetic material.

http://The Gene Editors Are Only Getting Started – School Information System https://apple.news/AgMnx20ynPYqkAuDX-kqR4A