Lindsay Clark in ComputerWeekly: “Open data, a movement which promises access to vast swaths of information held by public bodies, has started getting its hands dirty, or rather its feet.
Before a spade goes in the ground, construction and civil engineering projects face a great unknown: what is down there? In the UK, should someone discover anything of archaeological importance, a project can be halted – sometimes for months – while researchers study the site and remove artefacts….
During an open innovation day hosted by the Science and Technologies Facilities Council (STFC), open data services and technology firm Democrata proposed analytics could predict the likelihood of unearthing an archaeological find in any given location. This would help developers understand the likely risks to construction and would assist archaeologists in targeting digs more accurately. The idea was inspired by a presentation from the Archaeological Data Service in the UK at the event in June 2014.
The proposal won support from the STFC which, together with IBM, provided a nine-strong development team and access to the Hartree Centre’s supercomputer – a 131,000 core high-performance facility. For natural language processing of historic documents, the system uses two components of IBM’s Watson – the AI service which famously won the US TV quiz show Jeopardy. The system uses SPSS modelling software, the language R for algorithm development and Hadoop data repositories….
The proof of concept draws together data from the University of York’s archaeological data, the Department of the Environment, English Heritage, Scottish Natural Heritage, Ordnance Survey, Forestry Commission, Office for National Statistics, the Land Registry and others….The system analyses sets of indicators of archaeology, including historic population dispersal trends, specific geology, flora and fauna considerations, as well as proximity to a water source, a trail or road, standing stones and other archaeological sites. Earlier studies created a list of 45 indicators which was whittled down to seven for the proof of concept. The team used logistic regression to assess the relationship between input variables and come up with its prediction….”
Big Data, Machine Learning, and the Social Sciences: Fairness, Accountability, and Transparency
at Medium: “…So why, then, does granular, social data make people uncomfortable? Well, ultimately—and at the risk of stating the obvious—it’s because data of this sort brings up issues regarding ethics, privacy, bias, fairness, and inclusion. In turn, these issues make people uncomfortable because, at least as the popular narrative goes, these are new issues that fall outside the expertise of those those aggregating and analyzing big data. But the thing is, these issues aren’t actually new. Sure, they may be new to computer scientists and software engineers, but they’re not new to social scientists.
This is why I think the world of big data and those working in it — ranging from the machine learning researchers developing new analysis tools all the way up to the end-users and decision-makers in government and industry — can learn something from computational social science….
So, if technology companies and government organizations — the biggest players in the big data game — are going to take issues like bias, fairness, and inclusion seriously, they need to hire social scientists — the people with the best training in thinking about important societal issues. Moreover, it’s important that this hiring is done not just in a token, “hire one social scientist for every hundred computer scientists” kind of way, but in a serious, “creating interdisciplinary teams” kind of kind of way.
While preparing for my talk, I read an article by Moritz Hardt, entitled “How Big Data is Unfair.” In this article, Moritz notes that even in supposedly large data sets, there is always proportionally less data available about minorities. Moreover, statistical patterns that hold for the majority may be invalid for a given minority group. He gives, as an example, the task of classifying user names as “real” or “fake.” In one culture — comprising the majority of the training data — real names might be short and common, while in another they might be long and unique. As a result, the classic machine learning objective of “good performance on average,” may actually be detrimental to those in the minority group….
As an alternative, I would advocate prioritizing vital social questions over data availability — an approach more common in the social sciences. Moreover, if we’re prioritizing social questions, perhaps we should take this as an opportunity to prioritize those questions explicitly related to minorities and bias, fairness, and inclusion. Of course, putting questions first — especially questions about minorities, for whom there may not be much available data — means that we’ll need to go beyond standard convenience data sets and general-purpose “hammer” methods. Instead we’ll need to think hard about how best to instrument data aggregation and curation mechanisms that, when combined with precise, targeted models and tools, are capable of elucidating fine-grained, hard-to-see patterns….(More).”
Big video data could change how we do everything — from catching bad guys to tracking shoppers
Sean Varah at VentureBeat: “Everyone takes pictures and video with their devices. Parents record their kids’ soccer games, companies record employee training, police surveillance cameras at busy intersections run 24/7, and drones monitor pipelines in the desert.
With vast amounts of video growing vaster at a rate faster than the day before, and the hottest devices like drones decreasing in price and size until everyone has one (OK, not in their pocket quite yet) it’s time to start talking about mining this mass of valuable video data for useful purposes.
Julian Mann, the cofounder of Skybox Imaging — a company in the business of commercial satellite imagery and the developer advocate for Google Earth outreach — says that the new “Skybox for Good” program will provide “a constantly updated model of change of the entire planet” with the potential to “save lives, protect the environment, promote education, and positively impact humanity.”…
Mining video data through “man + machine” artificial intelligence is new technology in search of unsolved problems. Could this be the next chapter in the ever-evolving technology revolution?
For the past 50 years, satellite imagery has only been available to the U.S. intelligence community and those countries with technology to launch their own. Digital Globe was one of the first companies to make satellite imagery available commercially, and now Skybox and a few others have joined them. Drones are even newer, having been used by the U.S. military since the ‘90s for surveillance over battlefields or, in this age of counter-terrorism, playing the role of aerial detectives finding bad guys in the middle of nowhere. Before drones, the same tasks required thousands of troops on the ground, putting many young men and women in harm’s way. Today, hundreds of trained “eyes” safely located here in the U.S. watch hours of video from a single drone to assess current situations in countries far away….”
Smarter Than Us: The Rise of Machine Intelligence
Book by Stuart Armstrong at the Machine Intelligence Research Institute: “What happens when machines become smarter than humans? Forget lumbering Terminators. The power of an artificial intelligence (AI) comes from its intelligence, not physical strength and laser guns. Humans steer the future not because we’re the strongest or the fastest but because we’re the smartest. When machines become smarter than humans, we’ll be handing them the steering wheel. What promises—and perils—will these powerful machines present? Stuart Armstrong’s new book navigates these questions with clarity and wit.
Can we instruct AIs to steer the future as we desire? What goals should we program into them? It turns out this question is difficult to answer! Philosophers have tried for thousands of years to define an ideal world, but there remains no consensus. The prospect of goal-driven, smarter-than-human AI gives moral philosophy a new urgency. The future could be filled with joy, art, compassion, and beings living worthwhile and wonderful lives—but only if we’re able to precisely define what a “good” world is, and skilled enough to describe it perfectly to a computer program.
AIs, like computers, will do what we say—which is not necessarily what we mean. Such precision requires encoding the entire system of human values for an AI: explaining them to a mind that is alien to us, defining every ambiguous term, clarifying every edge case. Moreover, our values are fragile: in some cases, if we mis-define a single piece of the puzzle—say, consciousness—we end up with roughly 0% of the value we intended to reap, instead of 99% of the value.
Though an understanding of the problem is only beginning to spread, researchers from fields ranging from philosophy to computer science to economics are working together to conceive and test solutions. Are we up to the challenge?
A mathematician by training, Armstrong is a Research Fellow at the Future of Humanity Institute (FHI) at Oxford University. His research focuses on formal decision theory, the risks and possibilities of AI, the long term potential for intelligent life (and the difficulties of predicting this), and anthropic (self-locating) probability. Armstrong wrote Smarter Than Us at the request of the Machine Intelligence Research Institute, a non-profit organization studying the theoretical underpinnings of artificial superintelligence.”
Can we instruct AIs to steer the future as we desire? What goals should we program into them? It turns out this question is difficult to answer! Philosophers have tried for thousands of years to define an ideal world, but there remains no consensus. The prospect of goal-driven, smarter-than-human AI gives moral philosophy a new urgency. The future could be filled with joy, art, compassion, and beings living worthwhile and wonderful lives—but only if we’re able to precisely define what a “good” world is, and skilled enough to describe it perfectly to a computer program.
AIs, like computers, will do what we say—which is not necessarily what we mean. Such precision requires encoding the entire system of human values for an AI: explaining them to a mind that is alien to us, defining every ambiguous term, clarifying every edge case. Moreover, our values are fragile: in some cases, if we mis-define a single piece of the puzzle—say, consciousness—we end up with roughly 0% of the value we intended to reap, instead of 99% of the value.
Though an understanding of the problem is only beginning to spread, researchers from fields ranging from philosophy to computer science to economics are working together to conceive and test solutions. Are we up to the challenge?
A mathematician by training, Armstrong is a Research Fellow at the Future of Humanity Institute (FHI) at Oxford University. His research focuses on formal decision theory, the risks and possibilities of AI, the long term potential for intelligent life (and the difficulties of predicting this), and anthropic (self-locating) probability. Armstrong wrote Smarter Than Us at the request of the Machine Intelligence Research Institute, a non-profit organization studying the theoretical underpinnings of artificial superintelligence.”
Code of Conduct: Cyber Crowdsourcing for Good
Patrick Meier at iRevolution: “There is currently no unified code of conduct for digital crowdsourcing efforts in the development, humanitarian or human rights space. As such, we propose the following principles (displayed below) as a way to catalyze a conversation on these issues and to improve and/or expand this Code of Conduct as appropriate.
This initial draft was put together by Kate Chapman, Brooke Simons and myself. The link above points to this open, editable Google Doc. So please feel free to contribute your thoughts by inserting comments where appropriate. Thank you.
An organization that launches a digital crowdsourcing project must:
- Provide clear volunteer guidelines on how to participate in the project so that volunteers are able to contribute meaningfully.
- Test their crowdsourcing platform prior to any project or pilot to ensure that the system will not crash due to obvious bugs.
- Disclose the purpose of the project, exactly which entities will be using and/or have access to the resulting data, to what end exactly, over what period of time and what the expected impact of the project is likely to be.
- Disclose whether volunteer contributions to the project will or may be used as training data in subsequent machine learning research
- ….
An organization that launches a digital crowdsourcing project should:
- Share as much of the resulting data with volunteers as possible without violating data privacy or the principle of Do No Harm.
- Enable volunteers to opt out of having their tasks contribute to subsequent machine learning research. Provide digital volunteers with the option of having their contributions withheld from subsequent machine learning studies
- … “
When Experts Are a Waste of Money
Vivek Wadhwa at the Wall Street Journal: “Corporations have always relied on industry analysts, management consultants and in-house gurus for advice on strategy and competitiveness. Since these experts understand the products, markets and industry trends, they also get paid the big bucks.
But what experts do is analyze historical trends, extrapolate forward on a linear basis and protect the status quo — their field of expertise. And technologies are not progressing linearly anymore; they are advancing exponentially. Technology is advancing so rapidly that listening to people who just have domain knowledge and vested interests will put a company on the fastest path to failure. Experts are no longer the right people to turn to; they are a waste of money.
Just as the processing power of our computers doubles every 18 months, with prices falling and devices becoming smaller, fields such as medicine, robotics, artificial intelligence and synthetic biology are seeing accelerated change. Competition now comes from the places you least expect it to. The health-care industry, for example, is about to be disrupted by advances in sensors and artificial intelligence; lodging and transportation, by mobile apps; communications, by Wi-Fi and the Internet; and manufacturing, by robotics and 3-D printing.
To see the competition coming and develop strategies for survival, companies now need armies of people, not experts. The best knowledge comes from employees, customers and outside observers who aren’t constrained by their expertise or personal agendas. It is they who can best identify the new opportunities. The collective insight of large numbers of individuals is superior because of the diversity of ideas and breadth of knowledge that they bring. Companies need to learn from people with different skills and backgrounds — not from those confined to a department.
When used properly, crowdsourcing can be the most effective, least expensive way of solving problems.
Crowdsourcing can be as simple as asking employees to submit ideas via email or via online discussion boards, or it can assemble cross-disciplinary groups to exchange ideas and brainstorm. Internet platforms such as Zoho Connect, IdeaScale and GroupTie can facilitate group ideation by providing the ability to pose questions to a large number of people and having them discuss responses with each other.
Many of the ideas proposed by the crowd as well as the discussions will seem outlandish — especially if anonymity is allowed on discussion forums. And companies will surely hear things they won’t like. But this is exactly the input and out-of-the-box thinking that they need in order to survive and thrive in this era of exponential technologies….
Another way of harnessing the power of the crowd is to hold incentive competitions. These can solve problems, foster innovation and even create industries — just as the first XPRIZE did. Sponsored by the Ansari family, it offered a prize of $10 million to any team that could build a spacecraft capable of carrying three people to 100 kilometers above the earth’s surface, twice within two weeks. It was won by Burt Rutan in 2004, who launched a spacecraft called SpaceShipOne. Twenty-six teams, from seven countries, spent more than $100 million in competing. Since then, more than $1.5 billion has been invested in private space flight by companies such as Virgin Galactic, Armadillo Aerospace and Blue Origin, according to the XPRIZE Foundation….
Competitions needn’t be so grand. InnoCentive and HeroX, a spinoff from the XPRIZE Foundation, for example, allow prizes as small as a few thousand dollars for solving problems. A company or an individual can specify a problem and offer prizes for whoever comes up with the best idea to solve it. InnoCentive has already run thousands of public and inter-company competitions. The solutions they have crowdsourced have ranged from the development of biomarkers for Amyotrophic lateral sclerosis disease to dual-purpose solar lights for African villages….”
Training Students to Extract Value from Big Data
New report by the National Research Council: “As the availability of high-throughput data-collection technologies, such as information-sensing mobile devices, remote sensing, internet log records, and wireless sensor networks has grown, science, engineering, and business have rapidly transitioned from striving to develop information from scant data to a situation in which the challenge is now that the amount of information exceeds a human’s ability to examine, let alone absorb, it. Data sets are increasingly complex, and this potentially increases the problems associated with such concerns as missing information and other quality concerns, data heterogeneity, and differing data formats.
The nation’s ability to make use of data depends heavily on the availability of a workforce that is properly trained and ready to tackle high-need areas. Training students to be capable in exploiting big data requires experience with statistical analysis, machine learning, and computational infrastructure that permits the real problems associated with massive data to be revealed and, ultimately, addressed. Analysis of big data requires cross-disciplinary skills, including the ability to make modeling decisions while balancing trade-offs between optimization and approximation, all while being attentive to useful metrics and system robustness. To develop those skills in students, it is important to identify whom to teach, that is, the educational background, experience, and characteristics of a prospective data-science student; what to teach, that is, the technical and practical content that should be taught to the student; and how to teach, that is, the structure and organization of a data-science program.
Training Students to Extract Value from Big Data summarizes a workshop convened in April 2014 by the National Research Council’s Committee on Applied and Theoretical Statistics to explore how best to train students to use big data. The workshop explored the need for training and curricula and coursework that should be included. One impetus for the workshop was the current fragmented view of what is meant by analysis of big data, data analytics, or data science. New graduate programs are introduced regularly, and they have their own notions of what is meant by those terms and, most important, of what students need to know to be proficient in data-intensive work. This report provides a variety of perspectives about those elements and about their integration into courses and curricula…”
The nation’s ability to make use of data depends heavily on the availability of a workforce that is properly trained and ready to tackle high-need areas. Training students to be capable in exploiting big data requires experience with statistical analysis, machine learning, and computational infrastructure that permits the real problems associated with massive data to be revealed and, ultimately, addressed. Analysis of big data requires cross-disciplinary skills, including the ability to make modeling decisions while balancing trade-offs between optimization and approximation, all while being attentive to useful metrics and system robustness. To develop those skills in students, it is important to identify whom to teach, that is, the educational background, experience, and characteristics of a prospective data-science student; what to teach, that is, the technical and practical content that should be taught to the student; and how to teach, that is, the structure and organization of a data-science program.
Training Students to Extract Value from Big Data summarizes a workshop convened in April 2014 by the National Research Council’s Committee on Applied and Theoretical Statistics to explore how best to train students to use big data. The workshop explored the need for training and curricula and coursework that should be included. One impetus for the workshop was the current fragmented view of what is meant by analysis of big data, data analytics, or data science. New graduate programs are introduced regularly, and they have their own notions of what is meant by those terms and, most important, of what students need to know to be proficient in data-intensive work. This report provides a variety of perspectives about those elements and about their integration into courses and curricula…”
Big data is better data
Ted Talk by Kenneth Cukier, Data Editor of The Economist: “Self-driving cars were just the start. What’s the future of big data-driven technology and design? In a thrilling science talk, Kenneth Cukier looks at what’s next for machine learning — and human knowledge…”
Smarter video games, thanks to crowdsourcing
AAAS –Science Magazine: “Despite the stereotypes, any serious gamer knows it’s way more fun to play with real people than against the computer. Video game artificial intelligence, or AI, just isn’t very good; it’s slow, predictable, and generally stupid. All that stands to change, however, if GiantOtter, a Massachusetts-based startup, has its way, New Scientist reports. By crowdsourcing the AI’s learning, GiantOtter hopes to build systems where the computer can learn based on player’s previous behaviors, decision-making, and even voice communication—yes, the computer is listening in as you strategize. The hope is that by abandoning the traditional scripted programming models, AIs can be taught to mimic human behaviors, leading to more dynamic and challenging scenarios even in incredibly complex games like Blizzard Entertainment Inc.’s professionally played StarCraft II.“
Forget GMOs. The Future of Food Is Data—Mountains of It
Cade Metz at Wired: “… Led by Dan Zigmond—who previously served as chief data scientist for YouTube, then Google Maps—this ambitious project aims to accelerate the work of all the biochemists, food scientists, and chefs on the first floor, providing a computer-generated shortcut to what Hampton Creek sees as the future of food. “We’re looking at the whole process,” Zigmond says of his data team, “trying to figure out what it all means and make better predictions about what is going to happen next.”
The project highlights a movement, spreading through many industries, that seeks to supercharge research and development using the kind of data analysis and manipulation pioneered in the world of computer science, particularly at places like Google and Facebook. Several projects already are using such techniques to feed the development of new industrial materials and medicines. Others hope the latest data analytics and machine learning techniques can help diagnosis disease. “This kind of approach is going to allow a whole new type of scientific experimentation,” says Jeremy Howard, who as the president of Kaggle once oversaw the leading online community of data scientists and is now applying tricks of the data trade to healthcare as the founder of Enlitic.
Zigmond’s project is the first major effort to apply “big data” to the development of food, and though it’s only just getting started—with some experts questioning how effective it will be—it could spur additional research in the field. The company may license its database to others, and Hampton Creek founder and CEO Josh Tetrick says it may even open source the data, so to speak, freely sharing it with everyone. “We’ll see,” says Tetrick, a former college football linebacker who founded Hampton Creek after working on economic and social campaigns in Liberia and Kenya. “That would be in line with who we are as a company.”…
Initially, Zigmond and his team will model protein interactions on individual machines, using tools like the R programming language (a common means of crunching data) and machine learning algorithms much like those that recommend products on Amazon.com. As the database expands, they plan to arrange for much larger and more complex models that run across enormous clusters of computer servers, using the sort of sweeping data-analysis software systems employed by the likes of Google. “Even as we start to get into the tens and hundreds of thousands and millions of proteins,” Zigmond says, “it starts to be more than you can handle with traditional database techniques.”
In particular, Zigmond is exploring the use of deep learning, a form of artificial intelligence that goes beyond ordinary machine learning. Google is using deep learning to drive the speech recognition system in Android phones. Microsoft is using it to translate Skype calls from one language to another. Zigmond believes it can help model the creation of new foods….”
Zigmond’s project is the first major effort to apply “big data” to the development of food, and though it’s only just getting started—with some experts questioning how effective it will be—it could spur additional research in the field. The company may license its database to others, and Hampton Creek founder and CEO Josh Tetrick says it may even open source the data, so to speak, freely sharing it with everyone. “We’ll see,” says Tetrick, a former college football linebacker who founded Hampton Creek after working on economic and social campaigns in Liberia and Kenya. “That would be in line with who we are as a company.”…
Initially, Zigmond and his team will model protein interactions on individual machines, using tools like the R programming language (a common means of crunching data) and machine learning algorithms much like those that recommend products on Amazon.com. As the database expands, they plan to arrange for much larger and more complex models that run across enormous clusters of computer servers, using the sort of sweeping data-analysis software systems employed by the likes of Google. “Even as we start to get into the tens and hundreds of thousands and millions of proteins,” Zigmond says, “it starts to be more than you can handle with traditional database techniques.”
In particular, Zigmond is exploring the use of deep learning, a form of artificial intelligence that goes beyond ordinary machine learning. Google is using deep learning to drive the speech recognition system in Android phones. Microsoft is using it to translate Skype calls from one language to another. Zigmond believes it can help model the creation of new foods….”