sábado, 30 de noviembre de 2013

TellSpec: What's in your food?

A revolutionary hand-held device that tells you the allergens, chemicals, nutrients, calories, and ingredients in your food.

How does it work?
brings together laser spectroscopy, nanophotonics, and a unique mathematical algorithm in a revolutionary hand-held consumer device that can analyze the chemical composition of any food in less than 20 seconds.

The TellSpec handheld device beams a low-powered laser at the food you wish to analyze, measures the reflected light with a spectrometer, and sends the data via your smart phone, computer, or tablet to TellSpec’s servers in the cloud. Those servers use this data to deduce information about your food that is of interest to you. This information is then displayed on your computer, tablet or smart phone so you can intelligently decide if you want to buy or eat the food.


Open & Accessible Our platform is available to anyone, anywhere, to raise money for anything.

Free and Immediate You can start creating your campaign right now for free. There is no application process.

Open to Any Campaign You can raise money for anything, including for-profit ventures, creative ideas or personal needs.

Global Access No matter where you live, you can start your campaign and collect money from any country in the world as long as you have a valid bank account. Back to top Start Now ▶

Powerful Tools We have the tools you need to easily create, share and manage your campaign.

Create Use our platform to customize and publish a professional online funding campaign. No tech skills needed!
Share One-click social media integration, direct email and announcement features make it easy to spread the word, raise awareness and increase funding.

Track Manage contributions with our analytics tools, and stay on top of sending off your perks with our dashboard. Track who has claimed perks and how to get in touch with contributors to ensure they receive perks on time. Back to top Start Now ▶

Global Exposure Get merit-based promotion to make people around the world aware of your campaign.

Gogofactor Exposure Indiegogo promotes campaigns based on their gogofactor. This is a unique, merit-based algorithm that tracks the level of activity of each campaign based on how much you share, update, and attract funding. Work hard on your outreach, and we'll give you a boost!

Social Media and Community Outreach Based on your gogofactor activity, you may be eligible for exposure in our social media outreach, community and content creation, and press access. This is how Indiegogo helps to generate excitement and interest in your campaign! Back to top Start Now ▶

Personalization We provide campaign options and educational support that are right for you.

Your Money, Your Choice Keep all the money you raise, even if you don't meet your goal. Or, opt for Fixed Funding and only keep your money if your campaign reaches its goal.

Customizable Template You can customize your campaign page using our step-by-step set up. Write your story, upload videos, create a gallery of images, and offer perks.

Flexible Payment Options All campaigns can offer multiple funding types to their contributors, including PayPal and major credit cards.

Ongoing Education You can learn as you go from the experts (yes, real people!). Our Customer Happiness team and our blog posts will help you along the way - and we always respond in 24 hours or less.

TellSpec's CEO and CTO show a live demo of the TellSpec's food analysis algorithm.

A message from Dr Stephen Watson, TellSpec's CTO:

A message from Isabel Hoffmann, TellSpec's CEO:

TellSpec in the News:

ORIGINAL: Indiegogo

Fabian Oefner: Psychedelic science

By Fabian Oefner "Dancing Colors 08"
Fabian Oefner creates stunning visual representations of natural forces. 

Why you should listen to him: 
Fabian Oefner: Photographer. Creates stunning
visual representations of natural forces.

Fabian Oefner is a photographer and artist who wants to blend the disciplines of art and science. 
His psychedelic images capture natural phenomena and present them in unique and eye-catching ways. To date, subjects have included sound waves, iridescence, even magnetic ferroliquids and fire. His aim: to create images that appeal to both a viewer's heart and brain.

Oefner's photographs have been exhibited in various countries and are part of private collections around the globe. Besides pursuing his own projects, he also works on ad campaigns. He works and lives in Switzerland.

Oct 2013 

MIT discovers the location of memories: Individual neurons

MIT researchers have shown, for the first time ever, that memories are stored in specific brain cells. By triggering a small cluster of neurons, the researchers were able to force the subject to recall a specific memory. By removing these neurons, the subject would lose that memory.

As you can imagine, the trick here is activating individual neurons, which are incredibly small and not really the kind of thing you can attach electrodes to. To do this, the researchers used optogenetics, a bleeding edge sphere of science that involves the genetic manipulation of cells so that they’re sensitive to light. These modified cells are then triggered using lasers; you drill a hole through the subject’s skull and point the laser at a small cluster of neurons.

Now, just to temper your excitement, we should note that MIT’s subjects in this case are mice — but it’s very, very likely that the human brain functions in the same way. To perform this experiment, though, MIT had to breed genetically engineered mice with optogenetic neurons — and we’re a long, long way off breeding humans with optogenetic brains.

In the experiment, MIT gave mice an electric shock to create a fear memory in the hippocampus region of the brain (pictured above) — and then later, using laser light, activated the neurons where the memory was stored. The mice “quickly entered a defensive, immobile crouch,” strongly suggesting the fear memory was being recalled.

The main significance here is that we finally have proof that memories (engrams, in neuropsychology speak) are physical rather than conceptual. We now know that, as in Eternal Sunshine of the Spotless Mind, specific memories could be erased. It also gives us further insight into degenerative diseases and psychiatric disorders, which are mostly caused by the (faulty) interaction of neurons.The more we know about the moving pieces that make up our brains,” says Steve Ramirez, co-author of the paper. “The better equipped we are to figure out what happens when brain pieces break down.

Bear in mind, too, that this research follows on from MIT’s discovery last year of Npas4, the gene that controls the formation of memories; without Npas4, you cannot remember anything. MIT has successfully bred mice without the Npas4 gene.

The question now, though, is 
  • how memories are actually encodedcan we programmatically create new memories and thus learn entire subjects by inserting a laser into our brain
  • We know that a cluster of neurons firing can trigger the memory of your first kiss — but why
  • How can 100 (or 100,000) neurons, firing in a specific order, conjure up a beautifully detailed image of an elephant? 
We’ve already worked out how images are encoded by the optic nerve, so hopefully MIT isn’t too far away from finding out.

Read more at MIT or check out the research paper at Nature (paywalled)

ORIGINAL: Extreme Tech
March 23, 2012

martes, 26 de noviembre de 2013

La comida de insectos se industrializa y tiene mucho futuro por delante

El aviso de los organismos internacionales de alimentación queda claro: aunque existan más de mil millones de personas con exceso de peso y sobrealimentadas en el planeta, también existen otros mil millones que cada día se acuestan con hambre. Los alimentos en la actualidad están desequilibradamente repartidos, pero es que si la tendencia continúa así, según los estudios oficiales, la población planetaria no contará con viandas suficientes cuando alcancemos la cifra de 9.000 millones de personas, prevista para el 2050. Se buscan con urgencia soluciones para garantizar al mundo que todos tendremos acceso a una alimentación suficiente, segura y nutritiva.

Un equipo de estudiantes canadienses (Montreal) llamado Food Group Aspire, procedente de la universidad McGill, ha recibido el suculento premio Hult 2013 para poner en marcha su idea, cuyo fin es evitar esas previsibles hambrunas a las que ya se empieza a temer. Su objetivo: crear una industria global centrada en la elaboración de alimentos a partir de insectos, desde harina hasta bichos pasados a la parrilla. ¿Será este el destino de la cocina humana?

Add caption

Cuando nos presentamos al concurso vimos que el objetivo era ‘hacer frente a la inseguridad alimentaria’, cuenta Mohammed Ashour, líder y director del equipo. “Lo cierto es que después de hacer una lluvia de ideas solo llegábamos a soluciones aburridas y algo inútiles, así que decidimos buscar ideas locas, y un amigo nos habló de la posibilidad de comer insectos. Con un poco de investigación nos dimos cuenta de que muchas personas consumen insectos estacionalmente y que son nutritivos”.

En concreto, son más de 2.000 especies de insectos las que se comen en el planeta y son parte de la alimentación de dos millones y medio de personas, según Naciones Unidas (NU). El pasado enero un grupo de NU para la Agricultura y la Alimentación (FAO) formado por 60 expertos internacionales -en biología, nutrición y entomofagia- postularon que “comer insectos es bueno” porque son “son nutritivos, variados, económicos y hasta deliciosos”; y segundo, porque son una “vía para luchar contra el hambre” en este planeta de demografía en vertiginoso aumento.

Por ser los primeros en idear una industria formal con intención de desarrollar esa oportunidad, Food Group Aspire (compuesto por Ashour, Shobhita Soor, Zev Thompson, Jesse Pearlstein y Gabe Mott, un equipo multidisiplinar) fue el elegido en los Hult entre los 10.000 proyectos presentados.

Add caption

Con el cultivo, procesamiento y venta de insectos comestibles ofreceremos a las comunidades urbanas humildes un mejor acceso a una fuente eficiente y sostenible de proteínas y nutrientes”, cuenta el líder del grupo. ¿Suena loco? Lo es y no lo es”.

Nuestra intención es formalizar los mercados informales existentes y promover prácticas innovadoras de cultivo de insectos”, especifica Thompson, otro de los miembros. Su sistema consiste en enseñar a los habitantes de los países con mayor escasez alimentaria a procesar a gran escala la producción de insectos para crear harina, derivados de insectos, insectos asados y también sembrar la semilla que haga de esto un producto introducido formalmente en el mercado.

Para dar con el sistema perfecto, los universitarios visitaron países consumidores de bichos como México, Tailandia o Kenia, elaboraron diferentes técnicas de trabajo e incluso corrigieron errores de base. Inicialmente “queríamos que la elaboración de los productos de insectos se hiciera en los reductos urbanos marginales”, explica un miembro del conjunto, “pero después de nuestra investigación encontramos que la mayoría de la gente en los barrios pobres urbanos no quieren cultivar insectos porque tienen un espacio muy limitado y ensucian sus casas, así que nuestro modelo de negocio actual está destinado a agricultores semiurbanos y rurales que más tarde deberán distribuir sus productos en las ciudades”.

El alimento principal de Aspire es una harina fortificada, una mezcla de polvo de saltamontes (chapulines) con yuca, maíz o harina de trigo que tiene un sabor muy similar a la harina pura. Otros de sus productos son aperitivos o suplementos alimenticios. Y además sus avances también son útiles como pienso animal.

Los insectos son atrapados, lavados y escurridos”, explican el proceso. Después “se hornean y muelen”, así que son consumibles y totalmente higiénicos. Para dar el primer paso en este modelo de negocio y probar su eficacia, Aspire comenzará trabajando en Oaxaca (México) con las 10.000 toneladas de langostas que han podido adquirir gracias al premio. Una vez convertidas en harina, el producto se podrá emplear para hacer “pan, pasteles, sopas, guisos, salsas…”. También están analizando la posibilidad de desarrollar este proyecto piloto en Ghana a la vez que se realiza en México, donde el trabajo ya ha comenzado.

- ¿Cuánto dinero se ahorra al hacer una tonelada de harina de insectos en comparación con una tonelada de harina normal?

- No podemos responder con precisión ya que no hemos hecho harina a escala industrial todavía. Probablemente sea algo más cara.

-¿Y cuáles son las ventajas y desventajas de la primera en comparación con la segunda?

- No hay desventajas nutricionales para la harina fortificada (de insectos). Ventajas sí: un mayor contenido de minerales, particularmente el zinc , hierro, calcio, magnesio y mayor contenido de proteína.

-¿Creen que la gente de los países occidentales está dispuesta a comer insectos?

No tenemos ninguna ambición de vender cualquier cosa que no quieran en los países más desarrollados. Sin embargo, vale la pena señalar que cuando las personas de estos sitios comen grillos enteros asados, les suelen gustar mucho. ¿Por qué no iban a consumir la harina?

- En cuanto a la cantidad de materia prima, ¿cuántos kilos de harina se pueden obtener de las 10 toneladas de insectos?

-Hablamos de una proporción de 30 a 70 , lo que significa que 10 toneladas de insectos darían aproximadamente 33 toneladas de harina fortificada.

- Y en cuanto a sus beneficios como generador de empleo, ¿qué aporta vuestro proyecto?

- Esto depende en gran medida del insecto, pero en México prevemos que cada agricultor puede producir aproximadamente 2 kg de insectos en una sola bandeja cada mes. Un agricultor típico podría tener veinte contenedores, 40 kg por mes. Esto significa que 25 agricultores pueden producir hasta una tonelada de insectos.

- Oí que tuvieron algún problema de patentes con otro estudiante llamado Jakub Dzamba…

-Como este tema se está resolviendo aún, preferimos no comentar demasiado hasta que se acabe el proceso. Solo diremos que, igual que con Dzamba, consultamos con muchas personas para asesorarnos mientras trabajamos en el proyecto. Estamos agradecidos por sus asesoramientos y apoyos, y estaremos felices de poner el crédito de colaboraciones como la de Dzamba [este estudiante participó con unas ilustraciones que ya han sido retiradas del trabajo final], pero según lo confirmado por la organización del Premio Hult, las ilustraciones y el trabajo de diseño gráfico prestados por Jakub no eran pertinentes para los criterios de evaluación utilizados en la competencia, y por lo tanto, no era importantes en nuestra victoria. Le deseamos suerte, aunque a diferencia de él, para nosotros esto no es una cosa de firmar patentes, sino de resolver un problema mayor a nivel mundial.

- Parece que la industria alimentaria de los insectos a gran escala ya ha escuchado el pistoletazo de salida con el proyecto mexicano. ¿Próximas metas?

-Vemos una gran oportunidad de mercado en África Occidental en general, y además es el continente con mayor déficit alimentario de todos.

- Al final vuestra idea “loca” de aquella tormenta de ideas resultó no ser tan loca.

- Supongo que, para mí, los insectos comestibles hoy son como los vehículos eléctricos eran hace 10 años: cuando yo estaba empezando un negocio de bicicletas eléctricas con un amigo mío –responde Thompson -parecía muy experimental y vanguardista en su momento, pero ahora, diez años después, veo las bicicletas eléctricas en todas partes. Este proyecto de alimentación con insectos es nuestra oportunidad para potenciar la próxima generación y resolver algunos de los problemas más acuciantes del mundo. Esperamos que un día podamos mirar atrás y decir: estábamos en la vanguardia de los insectos comestibles en 2013.

Relacionado: ¿Acabaremos comiendo insectos para sobrevivir?

ORIGINAL: Yoroboku
by Jaled Abdelrahim(@jaledaa)
November 26, 2013

Updated: FDA Orders 23andMe to Stop Genetic Tests

The U.S. government is concerned that some of 23andMe’s health assessments could mislead customers.

On Friday, the U.S. Food and Drug Administration told 23andMe CEO Anne Wojcicki that her company “must immediately discontinue marketing the [Personal Genome Service].

Saliva Collection Kit Single
For $99, 23andMe will analyze the DNA in a saliva sample for genetic traits related to ancestry information, physical characteristics, disease risk, and drug response.
The health-related information is what concerns the FDA. The agency says that customers may make health decisions such as prophylactic breast removal surgery as a result of 23andMe’s report on their genetic risk for breast cancer. The problem, writes the FDA, is that the company has not provided the data to prove that the tests work, so consumers may make major health decisions based on faulty results.

Over a year ago, 23andMe announced that it was working with the FDA to get approval on at least some of the health traits in its service (see “Personal Genetics Company Seeks Regulatory Approval”). At the time, the company said it was looking forward to a collaborative process with the agency in the new territory of consumer genetics. So it was a bit surprising to read in the FDA’s letter that the company has not provided the agency with the data the regulatory agency requested in January to support marketing the health portions of the personal genomics test.

So what has 23andMe been doing instead? For one, they kicked off a large marketing campaign, which seems to have stirred up some ire in the FDA. From the agency’s letter to Wojcicki:

[The] FDA has not received any communication from 23andMe since May. Instead, we have become aware that you have initiated new marketing campaigns, including television commercials that, together with an increasing list of indications, show that you plan to expand the [Personal Genome Service’s] uses and consumer base without obtaining marketing authorization from FDA.

23andMe acknowledges that it has been remiss in responding to the FDA, but doesn’t say much else:

We recognize that we have not met the FDA’s expectations regarding timeline and communication regarding our submission. Our relationship with the FDA is extremely important to us and we are committed to fully engaging with them to address their concerns.

This post was updated 1:30 pm EST with comment from 23andMe.

ORIGINAL: Tech Review
Susan Young
November 25, 2013

Keep Yourself Warm In The Winter With Toasty Subway Heat

[Image: Tupungato via Shutterstock]
A new project in London is using the stifling hotness of a Tube tunnel for something better than making commuters sweaty: Saving on heating costs.

Underground train tunnels are quite hot. Your house in the winter is very cold. That problem and solution will soon meet in a London suburb's pioneering green energy project.

The project will capture waste heat from the London Underground Northern Line tunnel, as well as a nearby electrical substation, to supply more than 500 homes in the town of Islington with a greener heating supply. It’s expected to significantly lower home heating bills and help meet carbon dioxide emission reduction goals for the city. Record energy prices meant many families on fixed incomes spent the winter in misery, unsure whether to heat or eat.

It’s all part of the Council’s work to help people manage the rising cost of living. Last winter was one of the coldest for decades and record energy prices meant many families on fixed incomes spent it in misery, unsure whether to heat or eat,Richard Watts, leader of the Islington Council, said in a press release.

The project expands upon an existing green heating network in Islington that opened in November 2012 and already powers 700 homes. The network repurposes heat produced by the local power plant to heat homes, and contains 1.4 miles of pipes to deliver the heat to residences. Islington's council worked with London Mayor Boris Johnson and the local transit and power agencies to make the expansion to the Tube happen.

If it goes well, the idea of capturing waste heat from electricity substations that dot the London landscape could expand. UK Power Networks will use the experiment as part of a feasibility project. The mayor’s office, in its “Secondary Heat” report, also produced an assessment of many options for capturing and using waste heat locally, throughout London.

Would an idea like this work in New York City? That’s hard to say, though given the MTA’s ambitious sustainability goals, it’d be nice to see New Yorkers give it a try, so as not to be bested by their neighbors across the pond. As long as they guarantee that subway smell is not included in subway station heat.
ORIGINAL: FastCo Exist

World Map Installation uses E. Coli and Jellyfish Proteins to illuminate our population in 2100

A Buckminster Fuller-style Dymaxion Map
Terreform’s Bio City Map in full, one side
If you think about it, Buckminster Fuller’s Dymaxion Map is a perfect example of how reductive approaches to science may be necessary to resolve some of the world’s more pressing complications. To best understand the Earth as a total entity, Fuller suggested we go pre-Magellan, back to the days when the earth’s shape was physically unproven, by unravelling our beloved sphere to a flat, non-symmetrical surface. Like peeling an orange while keeping the peel intact.

Close up of Dymaxion Map
This is why it fits so well as a model for Terreform’s Bio City Lab – in ethos and in structure. Putting Fuller’s concept to practice, the New York-based design firm constructed a vertical plane of two-sided triangular pieces that model Earth’s surface, as if it were peeled directly off the mantle. Each side of the installation houses physical representations of data that snapshot a coming reality: by 2100, an anticipated 11 billion human bodies will be hustlin’ in all corners of the globe.

Instead of relying solely on computer algorithms or census trends, Terreform employs what it refers to as “bacteriography” to drive Bio City Lab’s glowing body. Strains of E. Coli and protein structures from sea anemones and jellyfish combine to bio-illuminate population fluctuations from now until 2100, ultimately mimicking the natural ebb-and-flow of urban densities with purely biological means.

Terreform’s website details why: “Bacteria in this constrained form and under the right conditions, behave almost identically to urban population patterns […] In many cases, they are as good as computational versions because they are the source which algorithms are derived from. In time, the mapping installation may illustrate patterns yet unobserved in typical digital models.

The protein structures are injected into the DNA of genetically modified E. Coli strains, which are then gathered in petri dishes and subjected to UV rays. These rays effectively flip a switch in the bacteria, resulting in a neon mesh of blues, greens, reds and yellows. Green glowing blotches indicate where we are now; red ones indicate what our numbers will look like in the coming century.

Opposite the petri dishes are mountainous 3D graphs detailing population peaks across 2100’s world.

As a result, the structure becomes both static and mutative: the rigid and plastic population graphs depict future projections, while the ongoing biological reactions depict the fluid, amorphous quality of population changes.

It also takes into account contemporary phenomena like megacities (urban areas with populations of more than 10 million) and instant cities (urban areas with an infrastructure erected in anticipation of a population, usually at the cusp of economic booms).

But instead of specifying which petri dishes or 3D graphs correlate with which cities, the Bio City Map is geographically indiscriminate. Current urban areas, countries, continents or even bodies of water remain unreferenced, so that the populationstatistics and data of each city come together to form a single, transcontinental urbanity. In turn, it becomes a city of cities.

Through this, the installation suggests that if we’re to tackle problems of saturated population density and their potential corollaries (water, energy, food, housing, etc. crises), we need to stop worrying about national or regional interest and look at the bigger global picture. Literally.

Bio City Map for Terreform’s Biological Urbanism at OCAD University, Toronto, Canada

Detail of population spike graph Terreform is an international contender when it comes to these things.
They’re one in a series of contemporary design firms looking to explore the romantic tendencies of futurism through experimental approaches to society building. Along with recent curations like Liam Young’s Future Perfect exhibition at the Lisbon Architecture Triennale (which we partially covered here) and the writings of William Meyers, they’re giving breath to the argument that creativity, technology, and biology must unite if we’re to effectively solve societal dilemmas down the road.

Each of these groups and creators recognize the need for cross-collaboration. It’s no longer just architects, just urbanists, or just engineers hashing out blueprints – it’s all of the above, plus a cadre of fiction authors, artists, futurists, mathematicians, and more. Which makes more than enough sense: how can you guide the growth of a society without soliciting the thoughts of those who grow its culture?

All photos courtesy of Terreform

ORIGINAL: Creators Project
By Johnny Magdaleno
Oct 22 2013

lunes, 25 de noviembre de 2013

Alien Squid Footage Surfaces

(ANIMAL NEWS/OCEANS DISCOVERY) Rare video footage (from 2007) of an unusual squid has the Internet in a frenzy over alien conspiracy theories. The squid in question is a 26-foot-long Magnapinna squid, one of the most curious-looking ocean dwellers.

The squid was recorded by a remotely-operated underwater vehicle (ROV) in the Gulf of Mexico. The footage shows the rarely seen mysterious squid displaying its floor-length tentacles. Read on to find out more about this fascinating alien-like creature and watch the footage in the video clip below. — Global Animal
The Magnapinna squid was recorded by a remotely operated
underwater vehicle in the Gulf of Mexico. Photo credit: Shell Oil

Daily Mail, Victoria Woollaston

Lurking deep beneath the Gulf of Mexico is a species of squid that wouldn’t look out of place in a sci-fi thriller.

The Mangapinna squid, sometimes referred to as the bigfin or long-arm squid, is around 26ft in length with thin elastic tentacles thought to be between 15 to 20 times larger than the squid’s body.

Adult bigfins have never been captured or sampled but rare video footage recorded by the Shell Oil company reveals their alien-like behaviour.

The footage was captured using a remotely operated underwater vehicle known as an ROV.

Shell Oil, along with other companies, uses the vehicles to study the water around its oil rigs and this particular recording was filmed in the Gulf of Mexico in the Perdido Area of Alaminos Canyon.

The rare sighting of the squid was discovered at a depth of more than 7,800 ft back in November 2007.

Shell oil has a rig located 200 miles off the coast of Houston, Texas.

Mangapinna squids were first discovered in 1907 but it wasn’t until 1988 that the first footage of the bizarre creatures were caught on camera by a submersible off the coast of Brazil.

Ten years later a Japanese submersible called Shinkai 6500 filmed another long-armed squid in the Indian Ocean south of Mauritius.

Photo credit: Shell Oil
The majority of other sightings have been in various canyons in the Gulf of Mexico.

A squid spotted in 2000 was thought to have been around 23ft long.

However, more recent sightings have estimated lengths in excess of 26ft.

Aside from their overall lengths, the arms and the tentacles of the Mangapinna squid are the same length and look identical.

Squids traditionally have two shorter arms and eight longer tentacles.
These ten appendages of the Mangapinna are also often held at right angles to the body, or mantle, which gives them the appearance of having elbows.

The arms and tentacles are said to stretch up to 20 times longer than the mantle while the fins are larger than other species and in some sightings were around 90 per cent as big.

It is thought the squids use their long arms to grab or trap food along the floor of the ocean, although this has never been seen in action.

Watch the squid footage in the video below.

ORIGINAL: Global Animal
By Sonia Horon 
November 21, 2013

Neural Networks and Deep Learning Book Project

A book that will teach you the core concepts of neural networks and deep learning

About the book
This book will teach you the core concepts of neural networks and deep learning. These are powerful machine learning techniques which have achieved outstanding results for problems in image recognition, speech recognition, and natural language processing. Neural networks and deep learning are now being adopted by many companies, including Google, Microsoft, and Facebook.

I'm writing this book to bridge the gap between popular accounts and the many technical papers on neural networks and deep learning. The book will make it easy and fun for people with programming and basic mathematical skills to come up to speed.

I love explaining complex technical subjects. I've written two previous books. The first book, "Quantum Computation and Quantum Information" (joint with Ike Chuang), is the standard text on quantum computing, and one of the ten most cited books in the history of physics. The second book, "Reinventing Discovery: The New Era of Networked Science", is a book for a general audience about networked science. It was named one of the best books of 2011 by The Financial Times and the Boston Globe.

In addition to my books I've written many technical articles, including "Lisp as the Maxwell's equations of software", "How to crawl a quarter billion webpages in 40 hours", and "Why Bloom filters work the way they do", all of which made the top five posts on Hacker News.

You can see a draft of chapter 1 of the book at neuralnetworksanddeeplearning.com.

The book will be made freely available online, under a Creative Commons Attribution-Non-Commercial license.

As an independent writer and scientist, the reason I'm undertaking this Indiegogo campaign is to give me some partial support while I complete the book.

Draft table of contents
  • Using neural nets to recognize handwritten digits: We get off to a flying start, creating a neural network that can solve a hard problem - recognizing handwritten digits.
  • Using backpropagation to speed up learning: We'll master the ins-and-outs of the backpropagation algorithm, which is the fundamental algorithm used to learn in neural nets, and the basis for deep learning.
  • Neural nets: the big picture: How do artificial neural nets compare to biological brains? Is there a simple universal algorithm for thinking? How can we use neural nets to solve problems in speech recognition and natural language processing? Can we use neural nets to compute an arbitrary function?
  • Deep learning: What makes deep neural networks hard to train with conventional approaches? How can we overcome those challenges? We'll see how deep neural nets can be pre-trained, and how they can learn high-level representations of knowledge from complex data.
  • Recent progress in image recognition: We'll dive into exciting recent work using deep learning to solve difficult problems in image recognition, including recognizing the images in ImageNet, and the Stanford-Google "cat neuron" paper.
  • The future of neural nets: Will neural nets help lead to artificial intelligence? Can they be used to simulate a human brain?
ORIGINAL: Indiegogo

Gordon Bell Prize Bubbles from Sequoia

Each year at SC, the ACM hands out one of the most coveted awards, the Gordon Bell Prize. The award, which became a regular feature of SC, began in 1987 and now carries a $10,000 prize sponsored by parallel computing luminary, Gordon Bell. Winners demonstrate high peak performance figures on real world applications or demonstrate other performance-geared achievements, including incredible advances in scaling, time to solution of scientific applications or other feats of HPC might.

This year the Gordon Bell award, chosen because of its demonstration of a high performance application went to “11 PFLOP/s Simulations of Cloud Cavitation Collapse,” by Diego Rossinelli, Babak Hejazialhosseini, Panagiotis Hadjidoukas and Petros Koumoutsakos, all of ETH Zurich, Costas Bekas and Alessandro Curioni of IBM Zurich Research Laboratory, and Steffen Schmidt and Nikolaus Adams of Technical University Munich.

The researchers, in collaboration with the Technical University of Munich and LLNL broke some serious computational fluid dynamics ground in their simulation, which maneuvered 6.4 million threads on the IBM Sequoia system. The simulation, according to IBM, stands as the “largest simulation ever in fluid dynamics by employing 13 trillion cells and reaching an unprecedented, for flow simulations, 14.4 petaflop sustained performance on Sequoia—73% of the supercomputer’s theoretical peak.

The bubble bursting exercises are more than just interesting to watch in action. These simulations model complex events related to clouds of collapsing bubbles, which can yield new insight in manufacturing, medicine and beyond as scientists seek to understand how they might “shatter” tumors, kidney stones or even fuel injection fluid interactions.

The researchers described their award-winning effort by pointing to how the “destructive power of cavitation reduces the lifetime of energy critical systems such as internal combustion engines and hydraulic turbines, yet it has been harnessed for water purification and kidney lithotripsy.” They go on to note that they were able to “advance by one order of magnitude the current state-of-the-art in terms of time to solution, and by two orders the geometrical complexity of the flow. The software successfully addresses the challenges that hinder the effective solution of complex flows on contemporary supercomputers, such as limited memory bandwidth, I/O bandwidth and storage capacity.

We were able to accomplish this using an array of pioneering hardware and software features within the IBM BlueGene/Q platform that allowed the fast development of ultra-scalable code which achieves an order of magnitude better performance than previous state-of-the-art,” said Alessandro Curioni, head of mathematical and computational sciences department at IBM Research – Zurich. “While the Top500 list will continue to generate global interest, the applications of these machines and how they are used to tackle some of the world’s most pressing human and business issues more accurately quantifies the evolution of supercomputing.

As IBM noted, These simulations are one to two orders of magnitude faster than any previously reported flow simulation. The last major achievement was earlier this year by a team at Stanford University which broke the one million core barrier, also on Sequoia.

This year the prize committee clarified their description of what it takes to render a winner, including the following criteria:

The prize winner is not selected simply on raw performance numbers. Rather, the Prize Committee seeks:
  • evidence of important algorithmic and/or implementation innovations
  • clear improvement over the previous state-of-the-art
  • solutions that don’t depend on one-of-a-kind architectures (systems that can only be used to address a narrow range of problems, or that can’t be replicated by others)
  • performance measurements that have been characterized in terms of scalability (strong as well as weak scaling), time to solution, efficiency (in using bottleneck resources, such as memory size or bandwidth, communications bandwidth, I/O), and/or peak performance
  • achievements that are generalizable, in the sense that other people can learn and benefit from the innovations
Solving an important scientific or engineering problem is important to demonstrate/justify the work, but scientific outcomes alone are not sufficient for this prize.

Nicole Hemsoth
November 22, 2013

sábado, 23 de noviembre de 2013

New York's Green Roofs Are Crawling With Fungi

Microscopic photo of an arbuscular mycorrhizal fungus courtesy of the USDA's Agricultural Research Service. USDA
Demand for green roofs might plummet if they became known as "fungal roofs." But that is what they are, at least in New York – and contrary to what it may sound like, it's not a bad thing.

The world just became a little more aware of the hidden-but-teeming biomass of green roofs thanks to the intrepid work of researchers from Barnard College, Columbia University, Fordham and the University of Colorado. Recently, these guys found themselves wondering if the gardens in the sky might support different kinds of life than the stuff at dog-pee level. It's a realm into which few scientific minds have tread. While green roofs as heat-island dampeners and rainwater-runoff plugs have been widely discussed, the extent to which they serve as urban "biodiversity reservoirs" (in the researchers' words) is something of a mystery.

So in the summer of 2011, the team set out to test the soil composition of 10 green roofs stationed at recreation centers throughout the five boroughs:
(Map courtesy of Jeremy Law at Columbia University)
Using soil corers, they hunted for fungi, because fungal communities play a key role in a roof garden's health and longevity. For comparison's sake, they also took samples from five city parks near some of the roofs, including Central Park and the High Line. A little magic from "[i]nductively coupled plasma atomic emission spectroscopy" at Alabama's Auburn University Soil Testing Laboratory, as well as a dollop of phospholipid fatty-acid extraction and Illumina-dye sequencing, and they had their results, which were published this month in the journal PLOS ONE.

So what were the conclusions? 
For one, these sun-kissed carpets of gray goldenrod and smooth blue aster are absolutely crawling with fungi. The researchers logged an average of 109 types of fungi per roof, such as Glomus, Acaulospora, Rhizophagus and Funneliformis, suggesting that green roofs can indeed contribute to urban biodiversity. As they explained:

We found that green roofs supported a diverse fungal community, with numerous taxa belonging to fungal groups capable of surviving in disturbed and polluted habitats. Across roofs, there was significant biogeographical clustering of fungal communities, indicating that community assembly of roof microbes across the greater New York City area is locally variable. Green roof fungal communities were compositionally distinct from city parks and only 54% of the green roof taxa were also found in the park soils.

In other words, the roofs are home to fungi not typically given to squelching around in normal parkland. They also seem to be better for growing stuff you might, you know, put in your mouth: While the soil in New York's parks showed a greater biomass of microbes, it also tested higher for heavy metals, a scourge of urban gardens that can be unhealthy if consumed in larger quantities.

Here's a comparison the researchers put together illustrating how the roofs stacked up against the parks, in terms of the abundance of fungal phyla:
Needless to say, this is hardly the first news of green roofs supporting life. The elevated gardens are routinely patrolled by insects and in some cases much larger fauna. In Australia, for instance, the Adelaide Zoo maintains several grassy roofs that are designed as homes for urban plants and wildlife, like reptiles, insects and bats.

And an immense green roof in the U.K., mounted on a wastewater treatment facility near Brighton, attracts seagulls and crows that pluck at its quaking grass in search of food. To fight those hungry birds, the roof's overseers have released even more animals over the roof – ferocious goshawks, a golden eagle and even a great horned owl.

ORIGINAL: Atlantic Cities
John Metcalfe
Mar 13, 2013

Ending Overfishing

Despite an increased awareness of overfishing, the majority of people still know very little about the scale of the destruction being wrought on the oceans. This film presents an unquestionable case for why overfishing needs to end and shows that there is still an opportunity for change. Reform of the EU's Common Fisheries Policy is almost complete. Fisheries ministers and members of the European Parliament, MEPs, are negotiating a deal for the future EU fisheries subsidies, which should support and end to EU overfishing. In the meantime you can support the campaign to end overfishing by, http://petition.bloomassociation.org


Stanford study could lead to paradigm shift in organic solar cell research

A new study by Stanford scientists overturns a widely held explanation for how organic photovoltaics turn sunlight into electricity.
Organic solar cells have long been touted as lightweight, low-cost alternatives to rigid solar panels made of silicon. Dramatic improvements in the efficiency of organic photovoltaics have been made in recent years, yet the fundamental question of how these devices convert sunlight into electricity is still hotly debated.

Now a Stanford University research team is weighing in on the controversy. Its findings, published in the Nov. 17 issue of the journal Nature Materials, indicate that the predominant working theory is incorrect and could steer future efforts to design materials that boost the performance of organic cells.

"We know that organic photovoltaics are very good," said study coauthor Michael McGehee, a professor of materials science and engineering at Stanford. "The question is, why are they so good? The answer is controversial."

A typical organic solar cell consists of two semiconducting layers made of plastic polymers and other flexible materials. The cell generates electricity by absorbing particles of light, or photons.
When the cell absorbs light, a photon knocks out an electron in a polymer atom, leaving behind an empty space, which scientists refer to as a hole. The electron and the hole immediately form a bonded pair called an exciton. The exciton splits, allowing the electron to move independently to a hole created by another absorbed photon. This continuous movement of electrons from hole to hole produces an electric current.

In the study, the Stanford team addressed a long-standing debate about what causes the exciton to split.

"To generate a current, you have to separate the electron and the hole," said senior author Alberto Salleo, an associate professor of materials science and engineering at Stanford. "That requires two different semiconducting materials. If the electron is attracted to material B more than material A, it drops into material B. In theory, the electron should remain bound to the hole even after it drops.

"The fundamental question that's been around a long time is, how does this bound state split?"
Some like it hot One explanation widely accepted by scientists is known as the "hot exciton effect." The idea is that the electron carries extra energy when it drops from material A to material B. That added energy gives the excited ("hot") electron enough velocity to escape from the hole.

But that hypothesis did not stand up to experimental tests, according to the Stanford team.

"In our study, we found that the hot exciton effect does not exist," Salleo said. "We measured optical emissions from the semiconducting materials and found that extra energy is not required to split an exciton."

So what actually causes electron-hole pairs to separate?
Stanford scientists may have resolved a debate about how organic solar cells turn sunlight into electricity. The question: What causes electron-hole pairs (excitons) to split apart? The likely answer: A gradient at the solar cell interface between disordered polymers and ordered buckyballs splits the exciton, allowing the electron (purple) to escape and produce an electric current. (Koen Vandewal / Stanford University)
"We haven't really answered that question yet," Salleo said. "We have a few hints. We think that the disordered arrangement of the plastic polymers in the semiconductor might help the electron get away."

In a recent study, Salleo discovered that disorder at the molecular level actually improves the performance of semiconducting polymers in solar cells. By focusing on the inherent disorder of plastic polymers, researchers could design new materials that draw electrons away from the solar cell interface where the two semiconducting layers meet, he said.

"In organic solar cells, the interface is always more disordered than the area farther away," Salleo explained. "That creates a natural gradient that sucks the electron from the disordered regions into the ordered regions. "
Improving energy efficiency The solar cells used in the experiment have an energy-conversion efficiency of about 9 percent. The Stanford team hopes to improve that performance by designing semiconductors that take advantage of the interplay between order and disorder.

"To make a better organic solar cell, people have been looking for materials that would give you a stronger hot exciton effect," Salleo said. "They should instead try to figure out how the electron gets away without it being hot. This idea is pretty controversial. It's a fundamental shift in the way people think about photocurrent generation."

Other authors of the paper are Koen Vandewal (lead author), Erik Hoke, William Mateker, Jason Bloking and George Burkhard of Stanford; Steve Albrecht, Marcel Schubert and Dieter Neher of the University of Potsdam; Johannes Widmer and Moritz Riede of the Institute for Applied Photophysics (IAPP); Jessica Douglas and Jean Frechet of the University of California-Berkeley; Aram Amassian of the King Abdullah University of Science and Technology (KAUST); and Alan Sellinger of the Colorado School of Mines and the University of Oxford. Author Kenneth Graham has a joint postdoctoral fellowship with Stanford and KAUST.

Support for the study was provided by the Stanford Center for Advanced Molecular Photovoltaics and the U.S. Department of Energy.

ORIGINAL: Stanford Engineering
By Mark Schwartz | Precourt Institute for Energy 
Tuesday, November 19, 2013

Mark Shwartz writes about energy technology for the Precourt Institute for Energy at Stanford University.
For more Stanford experts on energy and other topics, visit Stanford Experts.

The Impact of Brain and Mind Research

Understanding the brain is a grand challenge of science, and in April 2013, President Obama announced the federal BRAIN Initiative, whose goal is to create dramatic improvements in our understanding of brain function and dysfunction. Modeled loosely on the Human Genome Project, this initiative will require the development of new technologies, models, and computational approaches.

With so much at stake, what role can CMU and Pittsburgh play in this initiative? 
This panel of experts will discuss
  • the opportunities and challenges posed by the BRAIN Initiative, including 
  • the potential of this work to bring about revolutionary changes in our understanding of the brain; 
  • in our ability to understand, diagnose and treat brain disorders; and 
  • in the development of models that mimic brain functions.
Opening Remarks: Subra Suresh, Carnegie Mellon University

Moderator: Michael Tarr, Carnegie Mellon University Panelists:

Nathan Urban, Carnegie Mellon University
Marlene Behrmann, Carnegie Mellon University

Tom Mitchell, Carnegie Mellon University
Emery Brown, Massachusetts Institute of Technology, Harvard Medical School
Philip Rubin, Executive Office of the President of the United States
The Impact of Brain and Mind Research
Cèilidh Weekend, McConomy Auditorium, University Center, Carnegie Mellon University
September 28, 2013

miércoles, 20 de noviembre de 2013

The man who made robots smarter is now working on disrupting education

ORIGINAL: Venture Beat
Romain Nervil
November 20, 2013

Sebastian Thrun Wikipedia image
Sebastian Thrun has worked on making robots smarter, including “Stanley,” the driverless car that won the 2005 DARPA Challenge. Now he’s working on making humans smarter with Udacity, the online education company he co-founded in 2011.

Udacity offers affordable, accessible and relevant courses that serves hundreds of thousands of students from all over the world. Their curriculum, built together with industry leaders, bridges the gap between academia and skills needed for the 21st century tech workforce. Courses include modern programming, computer science, product design, and web development classes, as well as curriculum tracks in Data Science & Big Data.

Thrun will be speaking at the upcoming DataBeat/Data Science Summit, Dec. 4-5 in Redwood Shores, Calif., as part of a stellar lineup of data scientists, data analysts, and engineers building the next generation of Big Data tools.

So what does a robotics expert have to do with online education — and big data? 
A lot, as it turns out.

Online education might be an unusual outcome for a man whose early career was spent as a professor of computer science at Carnegie Mellon University and Stanford University in the 1990′s. He blended his two subjects of expertise, computer science and statistics, to help make robots smarter and more able to learn and interact with their environment. From 2003 on, he guided the team that built Stanley, helping give vehicle the necessary sensors to “see” the road, and software to make the right decisions and complete the race.

At the same time, he contributed to the machine learning algorithms that enabled the software to learn faster by incorporating human corrections, becoming better over time. Thrun went on the become a Google Fellow and publish a book on probabilistic programming techniques in robotics.

So what does that all have to do with Udacity? 
Why would a statistics and artificial intelligence guru start such a venture, if not to apply his machine learning chops to a grand challenge such as education?

Hundreds of thousands of students working on the classes give Udacity millions of interactions, generating gigabytes of progress data to mine in order to make the system smarter and more efficient.

In other words, the kind of probabilistic data analysis that helped make Stanley so smart could soon help make our educators smarter, teaching students more effectively and adapting to their needs more quickly, based on what they are actually doing in their classes.

And that, in turn, could make all of us smarter.

lunes, 18 de noviembre de 2013

El hombre que metió la evolución en un frasco

Richard Lenski (izqda) y Zachary Blount con las placas del experimento. (Imagen: Science)
En el laboratorio de Richard Lenski, en la Universidad Estatal de Michigan, hay seis frigoríficos que contienen 58.000 generaciones de bacterias. Su experimento comenzó en el año 1988 con doce cultivos idénticos de Escherichia coli y, después de 25 años, las bacterias han seguido reproduciéndose y evolucionando. En el mantenimiento de estos cultivos, y las siguientes generaciones, se han invertido más de 4 millones de dólares y han participado un centenar de personas que han alimentado y cuidado a estos microorganismos día y noche, fines de semana incluidos. Cada poco tiempo las bacterias se multiplican y se coloca otra generación de E. coli en una nueva placa, en condiciones idénticas a las anteriores. Se calcula que se han reproducido a un ritmo de 6,6 generaciones nuevas cada día, el equivalente - si nos reprodujéramos al mismo ritmo - a un millón de años de evolución humana.

¿Y con qué motivo desarrollaría un experimento tan largo y tan costoso? Para observar, en tiempo real, si la evolución se detiene en algún momento. Para entenderlo mejor, hay que tener en cuenta la primera idea de Lenski. Si mantenía durante el tiempo los doce cultivos de bacterias en condiciones idénticas, pensó, cada grupo se adaptaría a su entorno paulatinamente hasta alcanzar un nivel de adaptación óptimo a partir del cual no podrían mejorar y la evolución, por decirlo de alguna manera, habría llegado a un límite. Pero no podía estar más equivocado.

Datos del experimento. Ilustración: Science

Aparte de las bacterias que están reproduciéndose en tiempo real, en los estantes de los seis frigoríficos del laboratorio uno puede encontrar muestras de distintos momentos evolutivos de las bacterias (congelaban una placa cada 75 días). De este modo, uno puede escoger entre los 4.000 viales y retroceder hasta el punto del tiempo evolutivo que desee: a la generación 10.000, a la 20.000 o a la de hace solo un año. Y se puede saber qué grupo de E. coli ha conseguido mejores adaptaciones y está más "evolucionado" mediante una sencilla prueba: se descongela una muestra antigua, se pone al lado de una actual y se comprueba cuál se reproduce más rápido. Aquellas que se reproducen antes tendrán más opciones de sobrevivir, de modo que están mejor adaptadas. "Podemos poner a competir a organismos que vivieron en diferentes momentos del tiempo", asegura Lenski, "de modo que las bacterias evolucionadas pueden competir cara a cara con sus ancestros".

Lo que han observado los científicos es que, pese a las sospechas iniciales de Lenski, la evolución es un proceso imparable que nunca se detiene, incluso cuando el ambiente permanece inalterado. Durante los 25 años del experimento, la adaptación de las bacterias ha mejorado en una media del 70%. Las nuevas generaciones de bacterias se reproducen 1,7 veces en el tiempo en que las bacterias originales se reproducen una vez, cuentan en Science. Durante las primeras pruebas, recuerda Lenski en NPR, las bacterias doblaban su población en alrededor de una hora. Pasadas 50.000 generaciones, las bacterias doblan su número en apenas 40 minutos. Y los científicos creen que las futuras generaciones lo harán incluso más rápido. "En alrededor de un millón de años", sostiene Lenski, "el ritmo con el que se multiplican por dos estará en torno a los 20 minutos".

Dos cepas de bacterias E. coli compiten en una placa. Imagen: Michael Wiser.

El proceso, sin embargo, no es lineal. Las mejoras se producían al principio muy rápido y con el tiempo se han ralentizado, porque cada vez resulta más difícil mejorar. Además, algunas de las descendientes de las 12 cepas originales empezaron a progresar más rápido que otras y tomaron caminos evolutivos diferentes. Seis líneas, por ejemplo, desarrollaron un defecto en la reparación del ADN, pero en lugar de morir, el proceso dio lugar a un ritmo más rápido de mutaciones que el de sus compañeras. Hacia la generación 6.500, unos tres años después de empezar el experimento, dentro de uno de los frascos aparecieron dos tipos de E. coli diferentes: uno que formaba pequeñas colonias con células más pequeñas y otro que formaba colonias más grandes con bacterias más grandes. Los científicos esperaban que una acabara imponiéndose a la otra pero, para su sorpresa, ambas cepas han sobrevivido creando un ecosistema en el que la interacción entre ambas hace posible que ambas sigan adelante. "[Lenski] ha creado sus propias islas Galápagos", asegura Christopher Marx, microbiólogo de la Universidad de Harvard y uno de los investigadores que cuidó del experimento.

En otra ocasión, otra de las placas con E. coli se puso turbia debido a una acumulación inusualmente alta de bacterias y Lenski sospechó que se trataba de algún tipo de contaminación de la muestra. Tomaron del congelador un vial anterior de esa misma cepa y la reiniciaron desde otro punto, pero al cabo de tres semanas la placa se volvió a poner turbia. Cuando estudiaron la cepa Ara 3 - que así la habían bautizado - descubrieron que en lugar de alimentarse de glucosa, como las E. coli originales, habían evolucionado para alimentarse de citratos, un metabolismo que les permitía un ritmo de reproducción aún más alto.

"Este fue el suceso más importante de todo el experimento con E. coli", asegura el físico Christoph Adami. "Tener una nueva función compleja desarrollada aparentemente de la nada es muy destacable". Para ver qué había pasado volvieron a cultivar cepas anteriores y vieron que la adaptación para consumir citratos aparecía en 4 de los 72 cultivos. Finalmente, encontraron el minúsculo cambio genético que provocaba la aparición de las nuevas bacterias y que ponía encima de la mesa otra cuestión: cómo se forma una nueva especie. En biología, la definición de especie sigue sin estar muy clara, pero se considera que la prueba de fuego es que las dos especies no puedan tener descendencia común. Como esto no sucede con la reproducción asexual de las bacterias, la cosa se complica, y a lo más que han llegado es a mezclar los genomas y observar que se produce un nuevo tipo de bacteria menos adaptada.

En cualquier caso, como destacan en Science, el experimento de Lenski y las 12 cepas de E. coli ha servido para conocer mejor la evolución en tiempos más parecidos a los que ésta maneja: centenares de miles de generaciones. Estas primeras observaciones no son más que los primeros escarceos con el conocimiento de lo que sucede a nivel genético y un impulso, quizá, para que se pongan en marcha, y se mantengan, otros experimentos a largo plazo.

Referencias: The man who bottled evolution (Science) | Bacterial Competition In Lab Shows Evolution Never Stops (NPR)

Te puede interesar: En busca de las bacterias perdidas (Fogonazos)

ORIGINAL: Fogonazos
17 noviembre 2013