Curating the Internet: Science and technology micro-summaries for O...

Probabilistic computation as a cheaper alternative to quantum computing; A plan for advertisers to use blockchain to share data about campaigns; The vanishing birds story is over-hyped; Off-the-shelf stem cells could revolutionize medicine; and DARPA seeks direct brain-machine interaction

Fresh and Informative Content Daily: Welcome to my little corner of the blockchain

Straight from my RSS feed

Whatever gets my attention

Links and micro-summaries from my 1000+ daily headlines. I filter them so you don't have to.

pixabay license: source.

Who needs qubits? Factoring algorithm run on a probabilistic computer - Traditional computing architectures are fantastic for solving most types of problems, but there are some problems that they don't handle well. These include factoring numbers, optimization problems, and running neural networks. For these, and other problems, other architectures are being considered. For example, quantum computers can improve factoring; D-Wave has a quantum optimizer, which (unsurprisingly) is an architecture that improves optimization performance; and neural networks have been done with light, and with memristors. Now, a team of Japanese and American researchers is adding a new architecture to the fray. This architecture, probabilistic computing, lies somewhere between a quantum optimizer and a neural network computer, but it can be built with commercial-grade parts that run at room temperature. The fundamental unit of computation is called a p-bit, and it is just "likely to adopt a 0 or 1". Ironically, the concept was created by worsening the reliability of the highly reliable magnetic tunnel junction form of memory. At the moment, an 8 p-bit computer has used probabilistic algorithms to solve factoring problems for numbers up to 950, in times up to 15 seconds. This is not particularly impressive, but the advantage is that we know how to build large numbers of p-bits at relatively low cost, so the technique can be improved and turned loose on much larger numbers.

A big advertising agency is working on a blockchain-based plan to let walled gardens like Facebook share data with marketers - Advertisers are putting increasing pressure on platforms and publishers to tear down the walled gardens so that they can view the entirety of their advertising campaigns. So, "Jonathan Steuer, chief research officer at Omnicom Media Group, suggested during an Advertising Week panel that blockchain-like technology could help advertisers solve reach and frequency problems that happen when consumers are bombarded by too many ads." The tool would use varying levels of encryption and span across several datasets in order to minimize privacy concerns, and has been under development for six to nine months. Steuer said it would enable advertisers to create a standard framework to obtain things like reach, duplication, and attribution, without violating individual privacy, and it would be able to aggregate data across households. No mention of any privacy or security experts being involved in the framework design, though.

There Is No Impending Bird Apocalypse - A much hyped study that was led by Ken Rosenberg recently reported on the loss of nearly 3 billion birds in the United States during the years since 1970. This article quotes other experts to sort through the hype. For example, Brian Mcgill notes that 15% of the "lost birds" were actually not native North American species, and that wildlife managers have been actively trying to reduce the numbers of invasive species. McGill also notes that numbers for many species that thrive on farmland may have been artificially inflated in 1970 because of things like deforestation and prairie destruction. If so, the new numbers may indicate a return to a natural baseline. A third point from McGill is that most of the remaining lost birds come from the most populous species, so even where numbers are reduced, it often doesn't indicate a species-level problem. Todd Arnold adds that there are some threatened species, but that the findings have been overhyped and a more nuanced and accurate discussion is needed. It is noteworthy that the researchers engaged a professional communication team to manage publicity for this article when it was published. Rule of thumb: If a "news" story includes predictions like: catastrophe, crisis, apocalypse, ****ageddon, "tipping point", or collapse, then it's a fairly safe bet that someone is trying to manipulate you. h/t Real Clear Science

"Off-the-Shelf" Stem Cells Could Bring a Regenerative Revolution - After decades of research, scientists have learned techniques to coax a person's cells to grow back into stem cells, and then have the stem cells grow into cells of a particular type, for example: insulin-producing cells or specialized nerve cells. However, it is far too expensive and difficult to accomplish this at scale, in large part because every person's cells have particular quirks that make them unique. To get around this, scientists are working to build a generic sort of stem cell that can be grown at scale, and given to anyone without triggering an immune response. These cells, termed "hypoimmunogenic", are genetically altered to prevent them from triggering a response from the body's immune system. Recent progress in this direction has come from teams at the University of California San Francisco, the Harvard Stem Cell Institute, the Diabetes Research Institute, and the University of Louisville School of Medicine.

STEEM DARPA Seeks Brain-Machine Communication Systems for Active Service Robots - In this post, @rycharde begins by discussing the history of technologies for reading and interpreting signals inside the brain, and then reports on an effort by DARPA to cultivate brain-machine communications systems. The first technology for this purpose was the Electroencephalograph (EEG), which was first built in 1924 and first launched in a lab in 1936. This invention led to discoveries like brain waves and REM sleep. Next up is magnetic resonance imaging (MRI) which measures blood flow through various regions of the brain. And the final technology discussed is the intracranial Electroencephalograph (iEEG), which is an EEG where the sensors are intrusively placed inside of the skull, which bypasses the insulation that is provided by the skull. Many newer techniques, under development now, are extensions of the EEG technique. DARPA has contacted six teams to move forward with their research in this area, with the goal of developing 2-way communication between a brain and a machine. The methods they are exploring include the use of magnetic fields, a combination of optical and ultrasound, and semi-intrusive methods such as one that involves genetically engineering proteins in the brain to make the neurons more sensitive to magnetic fields. (A beneficiary setting of 10% has been applied to this post for @rycharde.)

In order to help make Steem the go to place for timely information on diverse topics, I invite you to discuss any of these links in the comments and/or your own response post.

Beneficiaries

Burn Steem/SBD - @null - 5%
Cited author(s) - @rycharde - 10%
Fundraising for the Rustin Golden Knights Marching Band - @rgkmb-unofficial - 10%
Posting and/or scheduling service (steempeak.com) - @steempeak - 5%
Science, Technology, Engineering, and Math (STEM) curation on Steem - @steemstem - 5%
Steem/API services (anyx.io) - anyx - 5%
Steem/RSS services (steemrss.com) - torrey.blog - 5%

My other open posts

(as of Monday afternoon)

@remlaps

@remlaps-lite

Fundraising for the Rustin Golden Knights Marching Band by @rgkmb-unofficial

About this series

Sharing a link does not imply endorsement or agreement, and I receive no incentives for sharing from any of the content creators.

Follow on steem: @remlaps-lite, @remlaps
If you are not on Steem yet, you can follow through RSS: remlaps-lite, remlaps.

Thanks to SteemRSS from philipkoon, doriitamar, and torrey.blog for the Steem RSS feeds!

Curating the Internet: Science and technology micro-summaries for October 1, 2019