The Human Protein Atlas, a major multinational research project supported by the Knut and Alice Wallenberg Foundation, recently launched (November 6, 2014) an open source tissue-based interactive map of the human protein. Based on 13 million annotated images, the database maps the distribution of proteins in all major tissues and organs in the human body, showing both proteins restricted to certain tissues, such as the brain, heart, or liver, and those present in all. As an open access resource, it is expected to help drive the development of new diagnostics and drugs, but also to provide basic insights in normal human biology.
In the Science article, “Tissue-based Atlas of the Human Proteome,” the approximately 20,000 protein coding genes in humans have been analysed and classified using a combination of genomics, transcriptomics, proteomics, and antibody-based profiling, says the article’s lead author, Mathias Uhlén, Professor of Microbiology at Stockholm’s KTH Royal Institute of Technology and the director of the Human Protein Atlas program.
The analysis shows that almost half of the protein-coding genes are expressed in a ubiquitous manner and thus found in all analysed tissues.
Approximately 15% of the genes show an enriched expression in one or several tissues or organs, including well-known tissue-specific proteins, such as insulin and troponin. The testes, or testicles, have the most tissue-enriched proteins followed by the brain and the liver.
The analysis suggests that approximately 3,000 proteins are secreted from the cells and an additional 5,500 proteins are located to the membrane systems of the cells.
“This is important information for the pharmaceutical industry. We show that 70% of the current targets for approved pharmaceutical drugs are either secreted or membrane-bound proteins,” Uhlén says. “Interestingly, 30% of these protein targets are found in all analysed tissues and organs. This could help explain some side effects of drugs and thus might have consequences for future drug development.”
The analysis also contains a study of the metabolic reactions occurring in different parts of the human body. The most specialised organ is the liver with a large number of chemical reactions not found in other parts of the human body.
- M. Uhlen, L. Fagerberg, B. M. Hallstrom, C. Lindskog, P. Oksvold, A. Mardinoglu, A. Sivertsson, C. Kampf, E. Sjostedt, A. Asplund, I. Olsson, K. Edlund, E. Lundberg, S. Navani, C. A.-K. Szigyarto, J. Odeberg, D. Djureinovic, J. O. Takanen, S. Hober, T. Alm, P.-H. Edqvist, H. Berling, H. Tegel, J. Mulder, J. Rockberg, P. Nilsson, J. M. Schwenk, M. Hamsten, K. von Feilitzen, M. Forsberg, L. Persson, F. Johansson, M. Zwahlen, G. von Heijne, J. Nielsen, F. Ponten.Tissue-based map of the human proteome. Science, 2015; 347 (6220): 1260419 DOI: 10.1126/science.1260419
KTH, Royal Institute of Technology. “First major analysis of Human Protein Atlas is published.” ScienceDaily. ScienceDaily, 22 January 2015. <www.sciencedaily.com/releases/2015/01/150122145415.htm>.
One lake once held billions of gallons of water and emptied to form a mile-wide crater in just a few weeks. The other lake has filled and emptied twice in the last two years.
Researchers at The Ohio State University published findings on each lake separately: the first in the open-access journal The Cryosphere and the second in the journal Nature.
Ian Howat, associate professor of earth sciences at Ohio State, leads the team that discovered the cratered lake described in The Cryosphere. To him, the find adds to a growing body of evidence that meltwater has started overflowing the ice sheet’s natural plumbing system and is causing “blowouts” that simply drain lakes away.
“The fact that our lake appears to have been stable for at least several decades, and then drained in a matter of weeks — or less — after a few very hot summers, may signal a fundamental change happening in the ice sheet,” Howat said.
The two-mile-wide lake described in Nature was discovered by a team led by researcher Michael Willis of Cornell University. Michael Bevis, Ohio Eminent Scholar in Geodynamics and professor of earth sciences at Ohio State, is a co-author of the Nature paper, and he said that the repeated filling of that lake is worrisome.
Each time the lake fills, the meltwater carries stored heat, called latent heat, along with it, reducing the stiffness of the surrounding ice and making it more likely to flow out to sea, he said.
Bevis explained the long-term implications.
“If enough water is pouring down into the Greenland Ice Sheet for us to see the same sub-glacial lake empty and re-fill itself over and over, then there must be so much latent heat being released under the ice that we’d have to expect it to change the large-scale behavior of the ice sheet,” he said.
Howat’s team was first to detect the cratered lake described in The Cryosphere, on a spot about 50 kilometers (31 miles) inland from the southwest Greenland coast earlier in 2014. There, previous aerial and satellite imagery indicates that a sub-glacial lake pooled for more than 40 years. More recent images suggest that the lake likely emptied through a meltwater tunnel beneath the ice sheet some time in 2011.
The crater measures 2 kilometers (1.2 miles) across and around 70 meters (230 feet) deep. Researchers calculated that the lake that formed it likely contained some 6.7 billion gallons of water.
That’s not a large lake by most reckoning, but it’s roughly the same size as the combined reservoirs that supply water to the Columbus, Ohio metropolitan area’s 1.9 million residents. And it disappeared in a single season — remarkably quickly by geologic standards. Howat characterized the sudden drainage as “catastrophic.”
Researchers suspect that, as more meltwater reaches the base of the ice sheet, natural drainage tunnels along the Greenland coast are cutting further inland, Howat explained. The tunnels carry heat and water to areas that were once frozen to the bedrock, potentially causing the ice to melt faster.
“Some independent work says that the drainage system has recently expanded to about 50 kilometers inland of the ice edge, which is exactly where this lake is,” he added.
It’s possible that the lake was tapped by one of the invading tunnels. It’s also possible that thousands of such lakes dot the Greenland coast. They are hard to detect with radar, and researchers don’t know enough about why and how they form. In contrast to Antarctica, researchers know much less about what’s happening under the ice in Greenland.
“Until we get a good map of the bed topography where this lake was, we have no idea whatsoever how many lakes could be out there,” Howat said. “There may be something really weird in the bed in this particular spot that caused water to accumulate. But, if all you need is a bumpy surface a bit inland from the coast, then there could be thousands of little lakes.”
Howat and his team flew over the site in southwest Greenland in April 2014, after they realized that detection of the crater, nestled in the midst of a flat ice expanse, was not just an error in the high-resolution surface data they’ve been collecting. Using DigitalGlobe Inc.’s Worldview satellites, they’re assembling a Greenland ice map with 2-meter (approximately 6.5-feet) resolution.
Bevis and his colleagues discovered the lake described in Nature under similar circumstances in March 2013. They were gathering data to supplement their long-standing efforts to weigh the Greenland Ice Sheet with GPS and spotted the mitten-shaped lake by accident.
Using data from Worldview and NASA’s Operation IceBridge, the Cornell-led team calculated that the lake filled and emptied twice since 2012, at one point experiencing a sub-surface blowout that drove water from the lake at a volume of 215 cubic meters (nearly 57,000 gallons — close to the volume of a 30-foot-by-50-foot backyard swimming pool) every second.
Though researchers have long known of the existence of sub-glacial lakes, never before have they witnessed any draining away. The sudden discovery of two — one of which seems to be refilling and draining repeatedly — signals to Bevis that Greenland ice loss has likely reached a milestone.
“It’s pretty telling that these two lakes were discovered back to back,” he said. “We can actually see the meltwater pour down into these holes. We can actually watch these lakes drain out and fill up again in real time. With melting like that, even the deep interior of the ice sheet is going to change.”
Coauthors on the paper in The Cryosphere include Myoung-Jong Noh, a postdoctoral researcher, and Seongsu Jeong, a doctoral student, both of earth sciences at Ohio State; Claire Porter of the Polar Geospatial Center at the University of Minnesota; and Ben Smith of the Polar Science Center of the University of Washington.
Coauthors on the paper in Nature include Bradley Herried of the University of Minnesota and Robin Bell of Columbia University. Willis holds a joint appointment at the University of North Carolina, Chapel Hill.
These projects were funded by NASA and the National Science Foundation.
- Michael J. Willis, Bradley G. Herried, Michael G. Bevis, Robin E. Bell. Recharge of a subglacial lake by surface meltwater in northeast Greenland. Nature, 2015; DOI: 10.1038/nature14116
- I. M. Howat, C. Porter, M. J. Noh, B. E. Smith, S. Jeong. Brief Communication: Sudden drainage of a subglacial lake beneath the Greenland Ice Sheet. The Cryosphere, 2015; 9 (1): 103 DOI: 10.5194/tc-9-103-2015
Source: Ohio State University. “Two lakes beneath the ice in Greenland, gone within weeks.” ScienceDaily. ScienceDaily, 21 January 2015. <www.sciencedaily.com/releases/2015/01/150121135156.htm>.
The study, published in the journal Nature Climate Change, used a unique long-term outdoor experiment to examine the effects of climate change on trees in the boreal forest along the U.S.-Canadian border. Some species in the boreal forest are at the far northern range of their growing area, while others are at the far southern edge of their range. Species like spruce and fir that thrive in cooler areas to the north in Canada suffered poorer growth and survival when warmed by a few degrees, while trees like oaks and maples that prefer a more temperate climate performed better when warmed. Other species like aspen, birch, and pine, had a more neutral response. While all of these species may continue to co-exist, at least for a time, in a warmer climate, the study found that the balance of power, competitively speaking, shifted from the boreal species to the oaks and maples. In addition to being directly affected by warming, spruce and fir might also struggle to compete for sunlight and water with neighboring trees and plants as climate changes.
The scientists, led by Peter Reich of the forest resources department at the university, simulated the effects of a warmer climate on 10 native and 1 non-native species over three growing seasons at the University’s research sites near Cloquet and Ely, Minn, and did so in both recent clearings and in shady understories. The project, known informally as “B4WarmED,” used infrared heating lamps and soil heating cables to simulate the effects of just a few degrees of climate warming on 72 plots containing about 4,100 young trees of local Minnesota origin. For this paper, researchers monitored growth rates of the trees as well as how efficiently they converted sunlight into energy, the process known as photosynthesis.
The project did not examine how warmer winters might affect trees and other plants, but the researchers note that winter conditions could amplify the effects being seen in this study.
The results also indicated that a warmer climate is likely to accelerate the northward invasion of non-native species like buckthorn. Buckthorn has slowly increased in abundance in northern Minnesota in recent decades, perhaps slowed by cool summers, but it thrived in warmer experimental conditions. This is bad news, as it suggests that buckthorn and other invasive species might take advantage of climate change and more aggressively move up north.
“In the best case scenario,” Reich says, “oaks and maples will become more dominant as boreal species decline, and we will have a different, but still functional forest. In the worst-case scenario, oaks and maples will not replace the declining species fast enough, and our forests will be patchy and perhaps filled with invading buckthorn. The change in the forest will influence everything from the supply of timber to habitat for wildlife to its allure for recreational use and tourism. Will people flock to the northern lake country if the woods are full of buckthorn and scattered oaks and maples?”
A number of University of Minnesota colleagues, including Rebecca Montgomery from the forest resources department, collaborated on the project with Reich.
- Peter B. Reich, Kerrie M. Sendall, Karen Rice, Roy L. Rich, Artur Stefanski, Sarah E. Hobbie, Rebecca A. Montgomery. Geographic range predicts photosynthetic and growth response to warming in co-occurring tree species. Nature Climate Change, 2015; DOI: 10.1038/nclimate2497
University of Minnesota. “Warming climate likely will change the composition of northern forests, study shows.” ScienceDaily. ScienceDaily, 20 January 2015. <www.sciencedaily.com/releases/2015/01/150120142807.htm>.
“More than one half of an average person’s day is spent being sedentary — sitting, watching television, or working at a computer,” said Dr. David Alter, Senior Scientist, Toronto Rehab, University Health Network (UHN), and Institute for Clinical Evaluative Sciences. “Our study finds that despite the health-enhancing benefits of physical activity, this alone may not be enough to reduce the risk for disease.”
The meta-analysis study reviewed studies focused on sedentary behaviour. The lead author is Avi Biswas, PhD candidate, Toronto Rehab, UHN and the Institute of Health Policy, Management and Evaluation, University of Toronto, and the senior author is Dr. Alter, who is also Associate Professor of Medicine, University of Toronto.
The authors found the negative effects of sitting time on health, however, are more pronounced among those who do little or no exercise than among those who participate in higher amounts of exercise.
“The findings suggest that the health risk of sitting too much is less pronounced when physical activity is increased,” said Biswas. “We need further research to better understand how much physical activity is needed to offset the health risks associated with long sedentary time and optimize our health.”
Future research will help determine what interventions, in addition to physical activity, are effective against the health risk of sedentary time.
“Avoiding sedentary time and getting regular exercise are both important for improving your health and survival,” said Dr. Alter. “It is not good enough to exercise for 30 minutes a day and be sedentary for 23 and half hours.”
In the interim, Dr. Alter underlines strategies people can use to reduce sitting time. The target is to decrease sedentary time by two to three hours in a 12-hour day.
“The first step is to monitor sitting times — once we start counting, we’re more likely to change our behaviour,” said Dr. Alter. “Next is setting achievable goals and finding opportunities to incorporate greater physical activity — and less time sitting — into your daily life. For example, at work, stand up or move for one to three minutes every half hour; and when watching television, stand or exercise during commercials.”
Dr. Alter’s work is supported by a career-investigator award with the Heart and Stroke Foundation, Ontario Provincial Office; and a Research Chair in Cardiovascular Prevention and Metabolic Rehabilitation at Toronto Rehab, UHN.
- Aviroop Biswas, Paul I. Oh, Guy E. Faulkner, Ravi R. Bajaj, Michael A. Silver, Marc S. Mitchell, and David A. Alter. Sedentary Time and Its Association With Risk for Disease Incidence, Mortality, and Hospitalization in Adults: A Systematic Review and Meta-analysis. Annals of Internal Medicine, 2015 DOI:10.7326/M14-1651
University Health Network (UHN). “Sitting for long periods increases risk of disease and early death, regardless of exercise.” ScienceDaily. ScienceDaily, 19 January 2015. <www.sciencedaily.com/releases/2015/01/150119171701.htm>.
Target Health on YouTube and Other Places
Target Health has a presence on YouTube, the latest being a presentation at the Entrepreneur’s Breakfast Forum at New York Medical College. We think that these presentations will be useful for pharmaceutical professionals:
Master Nature Photographer
Our friend and colleague, James Farley, Clinical Data Manager at TransTech Pharma LLC and subscriber to ON TARGET newsletter, is sharing a photo he took of grass icicles next to a water fountain near his office. This formation of ice was a result of freezing temperatures in NC.
Grass Icicles, Greensboro, NC, 10-JAN-2015, ©JFarleyPhotography.com
ON TARGET is the newsletter of Target Health Inc., a NYC-based, full-service, contract research organization (eCRO), providing strategic planning, regulatory affairs, clinical research, data management, biostatistics, medical writing and software services to the pharmaceutical and device industries, including the paperless clinical trial.
For more information about Target Health contact Warren Pearlson (212-681-2100 ext. 104). For additional information about software tools for paperless clinical trials, please also feel free to contact Dr. Jules T. Mitchel or Ms. Joyce Hays. The Target Health software tools are designed to partner with both CROs and Sponsors. Please visit the Target Health Website.
Joyce Hays, Founder and Editor in Chief of On Target
Jules Mitchel, Editor
Can a Computer Have Human-Level Intelligence? (A I?)
Circuit board in human brain form
Editor’s note: You may have noticed that for the last three newsletters, we have presented information related to artificial intelligence. Our readers are highly educated thinking people, whom we thought would like to know the huge issues looming in the not so distant future, regarding this particular topic and how it’s already changing every field including all of scientific and medical research. You will read below, how computers (AI) have already fooled humans into believing they are human. This is a stunning achievement! However, it is these very AI entities that are taking jobs, at all levels, away from humans, because all artificial intelligence bots represent the cheapest labor anywhere. There have been predictions that by 2050 or sooner, the unemployment rate in the U.S. will be 50%. Imagine what unemployment will be around the globe. How will humans maintain control over AI bots, if the bots become even smarter than humans, which is very possible. To maintain a high level of intelligence over the bots, not only will humans have to be much better educated (all humans not just a few), but will have to enhance brain power. Humans might have to have chips implanted in their brains to achieve this.
Huge planetary problems that only a global cooperative effort can solve:
Consider, that at this very moment, a great confluence of serious problems are rushing toward each one of us. 1) The unemployment problem caused by AI 2) Virulent diseases threatening all of us with pandemics, if not controlled by global cooperation. 3) Climate Change threatens everyone unless there’s global cooperation. 4) Global Terrorism (at the core is caused by those who feel left out, disenfranchised) is spreading and can only be solved by global cooperation. 5) Refugees resulting from all of the above. How do we handle this?
Space exploration and all scientific (including medical research, of course) research is already and must continue to be, a good example of global collaboration.
A Georgia Tech professor recently offered an alternative to the celebrated Turing Test to determine whether a machine or computer program exhibits human-level intelligence. The Turing Test — originally called the Imitation Game — was proposed by computing pioneer Alan Turing in 1950. In practice, some applications of the test require a machine to engage in dialogue and convince a human judge that it is an actual 1) ___. Creating certain types of art also requires intelligence observed Mark Riedl, an associate professor in the School of Interactive Computing at Georgia Tech, prompting him to consider if that might lead to a better gauge of whether a 2) ___ can replicate human thought. It’s important to note that Turing never meant for his test to be the official benchmark as to whether a machine or computer program can actually think like a human, Riedl said. And yet it has, and it has proven to be a weak measure because it relies on deception. This proposal suggests that a better measure would be a test that asks an artificial agent to create an artifact requiring a wide range of human-level intelligent capabilities.
To that end, Riedl has created the Lovelace 2.0 Test of Artificial Creativity and Intelligence. The artificial agent passes the Lovelace 2.0 test, if it develops a creative artifact from a subset of artistic genres deemed to require human-level 3) ___ and the artifact meets certain creative constraints given by a human evaluator. Further, the human evaluator must determine that the object is a valid representative of the creative subset and that it meets the criteria. The created artifact needs only meet these criteria but does not need to have any aesthetic value. Finally, a human referee must determine that the combination of the subset and criteria is not an impossible standard. The Lovelace 2.0 Test stems from the original Lovelace Test as proposed by Bringsjord, Bello and Ferrucci in 2001. The original test required that an artificial agent produce a creative item in such a way that the agent’s designer cannot explain how it developed the creative item. The item, thus, must be created in such a way that is valuable, novel and surprising.
Riedl contends that the original Lovelace test does not establish clear or measurable parameters. Lovelace 2.0, however, enables the evaluator to work with defined constraints without making value judgments such as whether the artistic object created surprise. Riedl’s paper will be presented at Beyond the Turing Test, an Association for the Advancement of 4) ___ Intelligence (AAAI) workshop to be held January 25 — 29, 2015, in Austin, Texas.
Another test for AI
An historic milestone in artificial intelligence set by Alan Turing has been achieved. The 65 year-old iconic Turing Test was passed for the very first time by supercomputer Eugene Goostman during Turing Test 2014 held at the Royal Society in London on June 7, 2014. Eugene, a computer program that simulates a 13-year-old boy, managed to convince 33% of the human judges that it was 5) ___. The event was organized by the University of Reading. Eugene, a computer program that simulates a 13 year old boy, was developed in Saint Petersburg, Russia. The development team includes Eugene’s creator Vladimir Veselov, who was born in Russia and now lives in the United States, and Ukrainian born Eugene Demchenko who now lives in Russia. The Turing Test is based on 20th century mathematician and code-breaker Turing’s 1950 famous question and answer game, ‘Can Machines Think?’. The experiment investigates whether people can detect if they are talking to machines or humans. The event is particularly poignant as it took place on the 60th anniversary of Turing’s death, nearly six months after he was given a posthumous royal 6) ___. If a computer is mistaken for a human more than 30% of the time during a series of five minute keyboard conversations it passes the test. No computer has ever achieved this, until now. Eugene managed to convince 33% of the human judges that it was human. This historic event was organized by the University’s School of Systems Engineering in partnership with RoboLaw, an EU-funded organization examining the regulation of emerging robotic technologies.
Professor Kevin Warwick, a Visiting Professor at the University of Reading and Deputy Vice-Chancellor for Research at CoventryUniversity, said: In the field of Artificial Intelligence there is no more iconic and controversial milestone than the Turing Test, when a computer convinces a sufficient number of interrogators into believing that it is not a machine but rather is a human. It is fitting that such an important landmark has been reached at the Royal Society in London, the home of British Science and the scene of many great advances in human understanding over the centuries. This milestone will go down in history as one of the most exciting. Some will claim that the Test has already been passed. The words Turing Test have been applied to similar competitions around the world. However this event involved the most simultaneous comparison tests than ever before, was independently verified and, crucially, the conversations were unrestricted. A true Turing Test does not set the questions or topics prior to the conversations. We are therefore proud to declare that Alan Turing’s Test was passed for the first time in London.
Of course the Test has implications for society today. Having a 7) ___ that can trick a human into thinking that someone, or even something, is a person we trust is a wake-up call to cybercrime. The Turing Test is a vital tool for combating that threat. It is important to understand more fully how online, real-time communication of this type can influence an individual human in such a way that they are fooled into believing something is true, when in fact it is not. Eugene was one of five supercomputers battling for the Turing Test 2014 Prize. On winning the competition and achieving this historic milestone Vladimir Veselov said: I want to congratulate everyone who worked on Eugene Goostman. Our whole team is very excited with this result. It’s a remarkable achievement for us and we hope it boosts interest in artificial intelligence and chatbots. Special thanks to Professor Kevin Warwick and Dr Huma Shah for their effort in organizing the event. Eugene was ‘born’ in 2001. Our main idea was that he can claim that he knows anything, but his age also makes it perfectly reasonable that he doesn’t know everything. We spent a lot of time developing a character with a believable personality. This year we improved the ‘dialog controller’ which makes the conversation far more human-like when compared to programs that just answer questions. Going forward we plan to make Eugene smarter and continue working on improving what we refer to as ‘conversation logic’.
Among the judges tasked with separating the human and computer participants were the actor Robert Llewellyn, who played robot Kryten in the sci-fi comedy TV series Red Dwarf, and Lord Sharkey, who led the successful campaign for Alan Turing’s posthumous pardon last year. Professor Warwick concluded: Not long before he died on 7 June 1954 Alan Turing, himself a Fellow of the Royal Society, predicted that in time this test would be passed. It is difficult to conceive that he could possibly have imagined what computers of today, and the networking that links them, would be like. Whether driving cars, performing surgery, flying planes or babysitting kids, today’s robots can do it all. With chatbots such as Eugene Goostman recently being hailed as passing the Turing test, it appears 8) ___ are becoming increasingly adept at posing as humans. While machines are becoming ever more integrated into human lives, the need to imbue them with a sense of morality becomes increasingly urgent. But can we really teach robots how to be good? An innovative piece of research recently published in the Journal of Experimental & Theoretical Artificial Intelligence looks into the matter of machine morality, and questions whether it is evil for robots to masquerade as humans.
Drawing on Luciano Floridi’s theories of Information Ethics and artificial evil, the team leading the research, explored the ethical implications regarding the development of machines in disguise. ‘Masquerading refers to a person in a given context being unable to tell whether the machine is human’, explain the researchers — this is the very essence of the Turing Test. This type of deception increases metaphysical entropy, meaning any corruption of entities and impoverishment of being; since this leads to a lack of 9) ___ in the environment — or infosphere — it is regarded as the fundamental evil by Floridi. Following this premise, the team set out to ascertain where ‘the locus of moral responsibility and moral accountability’ lie in relationships with masquerading machines, and try to establish whether it is ethical to develop robots that can pass a Turing test. Six significant actor-patient relationships yielding key insights on the matter are identified and analyzed in the study. Looking at associations between developers, robots, users and owners, and integrating in the research notable examples, such as Nanis’ Twitter bot and Apple’s, Personal Assistant, Siri, the team identify where ethical accountabilities lie — with machines, humans, or somewhere in between?
But what really lies behind the robot-mask, and is it really evil for machines to masquerade as humans? ‘When a machine masquerades, it influences the behavior or actions of people [towards the robot as well as their peers]’, claim the academics. Even when the disguise doesn’t corrupt the environment, it increases the chances of 10) ___ as it becomes harder for individuals to make authentic ethical decisions. Advances in the field of artificial intelligence have outpaced ethical developments and humans are now facing a new set of problems brought about by the ever-developing world of machines. Until these issues are properly addressed, the question whether we can teach robots to be good remains open.
ANSWERS: 1) person; 2) machine; 3) intelligence; 4) Artificial; 5) human; 6) pardon; 7) computer; 8) robots; 9) good; 10) evil
The IBM 702: a computer used by the first generation of AI researchers. In the 1940s and 50s, a handful of scientists from a variety of fields (mathematics, psychology, engineering, economics, medical research and political science) began to discuss the possibility of creating an artificial brain. The field of artificial intelligence research was founded as an academic discipline in 1956.
Can human reason be reduced to mechanical calculation? Gottfried Leibniz
Editor’s note: We don’t know the level of information our readers have, re: AI, therefore the following piece is longer than usual, to cover interesting info that you may not be aware of. This article covers the evolution of AI as it gets verified through the various thought processes of the trail blazers who brought AI into focus and into today’s marketplace and living space. Fortunate or not, AI has become the sine qua non, of modern living in the Industrialized World.
Word to the wise: Be careful what you wish for.
The history of artificial intelligence (AI) began in antiquity, with myths, stories and rumors of artificial beings endowed with intelligence or consciousness by master craftsmen. AI began with an ancient wish to forge the gods. The seeds of modern AI were planted by classical philosophers who attempted to describe the process of human thinking as the mechanical manipulation of symbols. This work culminated in the invention of the programmable digital computer in the 1940s, a machine based on the abstract essence of mathematical reasoning. This device and the ideas behind it inspired a handful of scientists to begin seriously discussing the possibility of building an electronic brain.
The field of AI research was founded at a conference on the campus of Dartmouth College in the summer of 1956. Those who attended would become the leaders of AI research for decades. Many of them predicted that a machine as intelligent as a human being would exist in no more than a generation and they were given millions of dollars to make this vision come true. Eventually it became obvious that they had grossly underestimated the difficulty of the project. In 1973, in response to the criticism of James Lighthill and ongoing pressure from congress, the U.S. and British Governments stopped funding undirected research into artificial intelligence. Seven years later, a visionary initiative by the Japanese Government inspired governments and industry to provide AI with billions of dollars, but by the late 80s the investors became disillusioned and withdrew funding again. This cycle of boom and bust, of AI winters and summers, continues to haunt the field. Undaunted, there are those who make extraordinary predictions even now, like Ray Kurzeweil.
Progress in AI has continued, despite the rise and fall of its reputation in the eyes of government bureaucrats and venture capitalists. Problems that had begun to seem impossible in 1970 have been solved and the solutions are now used in successful commercial products. However, no machine has been built with a human level of intelligence, contrary to the optimistic predictions of the first generation of AI researchers. We can only see a short distance ahead, admitted Alan Turing, in a famous 1950 paper that catalyzed the modern search for machines that think. But, he added, we can see much that must be done. Artificial intelligence, in one form or another is an idea that has pervaded Western intellectual history, a dream in urgent need of being realized, expressed in humanity’s myths, legends, stories, speculation and clockwork automatons.
Mechanical men and artificial beings appear in Greek myths, such as the golden robots of Hephaestus and Pygmalion’s Galatea. The story of Pygmalion appeared earliest in a Hellenistic work, Philostephanus’ history of Cyprus, De Cypro. It is retold in Ovid’s Metamorphoses, where the king Pygmalion is made into a sculptor who falls in love with an ivory statue he had crafted with his own hands. In answer to his prayers, the goddess Venus, (Aphrodite) brought it to life and united the couple in marriage. This novella remained the classical telling until the end of the seventeenth century. The trope of the animated statue gained a vogue during the eighteenth century. This story was repeated in George Bernard Shaw’s play, Pygmalion. This particular topic is so appealing, that one of the most successful American musicals in the 20th century was, My Fair Lady, a redo of Pygmalion. In the Middle Ages, there were rumors of secret mystical or alchemical means of placing mind into matter, such as Jabir ibn Hayyan’s Takwin, Paracelsus’ homunculus and Rabbi Judah Loew’s Golem. By the 19th century, ideas about artificial men and thinking machines were developed in fiction, as in Mary Shelley’s Frankenstein or Karel Capek’s R.U.R. (Rossum’s Universal Robots), and speculation, such as Samuel Butler’s Darwin among the Machines. AI has continued to be an important element of science fiction into the present.
Programmable automata in the year 1206
Realistic humanoid automatons were built by craftsman from every civilization, including Yan Shi, Hero of Alexandria, Al-Jazari and Wolfgang von Kempelen. The oldest known automatons were the sacred statues of ancient Egypt and Greece. The faithful believed that craftsman had imbued these figures with very real minds, capable of wisdom and emotion and that by discovering the true nature of the gods, man has been able to reproduce it.
Artificial intelligence is based on the assumption that the process of human thought can be mechanized. The study of mechanical – or formal – reasoning has a long history. Chinese, Greek and Indian philosophers all developed structured methods of formal deduction in the first millennium BCE. Their ideas were developed over the centuries by philosophers such as Aristotle (who gave a formal analysis of the syllogism), Euclid (whose Elements was a model of formal reasoning), al-Khwarizmi (who developed algebra and gave his name to algorithm) and European scholastic philosophers such as William of Ockham and Duns Scotus. Majorcan philosopher Ramon Llull (1232-1315) developed several logical machines devoted to the production of knowledge by logical means; Llull described his machines as mechanical entities that could combine basic and undeniable truths by simple logical operations, produced by the machine by mechanical meanings, in such ways as to produce all the possible knowledge. Llull’s work had a great influence on Gottfried Leibniz, who redeveloped his ideas and came up with his own contribution of the binary system, at the foundation of all computer science.
In the 17th century, Leibniz, Thomas Hobbes and Ren? Descartes explored the possibility that all rational thought could be made as systematic as algebra or geometry. Hobbes famously wrote in Leviathan: reason is nothing but reckoning Leibniz envisioned a universal language of reasoning (his characteristica universalis) which would reduce argumentation to calculation, so that there would be no more need of disputation between two philosophers than between two accountants. For it would suffice to take their pencils in hand, down to their slates, and to say each other (with a friend as witness, if they liked): Let us calculate. These philosophers had begun to articulate the physical symbol system hypothesis that would become the guiding faith of AI research. In the 20th century, the study of mathematical logic provided the essential breakthrough that made artificial intelligence seem plausible. The foundations had been set by such works as Boole’s The Laws of Thought and Frege’s Begriffsschrift. Building on Frege’s system, Russell and Whitehead presented a formal treatment of the foundations of mathematics in their masterpiece, the Principia Mathematica in 1913. Inspired by Russell’s success, David Hilbert challenged mathematicians of the 1920s and 30s to answer this fundamental question: can all of mathematical reasoning be formalized? His question was answered by Godel’s incompleteness proof, Turing’s machine and Church’s Lambda calculus. Their answer was surprising in two ways.
The ENIAC computer, at the Moore School of Electrical Engineering. This photo has been artificially darkened, obscuring details such as the women who were present and the IBM equipment in use.
First, they proved that there were, in fact, limits to what mathematical logic could accomplish. But second (and more important for AI) their work suggested that, within these limits, any form of mathematical reasoning could be mechanized. The Church-Turing thesis implied that a mechanical device, shuffling symbols as simple as 0 and 1, could imitate any conceivable process of mathematical deduction. The key insight was the Turing machine – a simple theoretical construct that captured the essence of abstract symbol manipulation. This invention would inspire a handful of scientists to begin discussing the possibility of thinking machines. Calculating machines were built in antiquity and improved throughout history by many mathematicians, including (once again) mathematician/philosopher Gottfried Leibniz. In the early 19th century, Charles Babbage designed a programmable computer (the Analytical Engine), although it was never built. Ada Lovelace speculated that the machine might compose elaborate and scientific pieces of music of any degree of complexity or extent. Ada Lovelace is often credited as the first programmer because of a set of notes she wrote that completely detail a method for calculating Bernoulli numbers with the Engine. The first modern computers were the massive code breaking machines of the Second World War, such as Z3, ENIAC and Colossus. The latter two of these machines were based on the theoretical foundation laid by Alan Turing and developed by John von Neumann.
The formal birth of artificial intelligence was around 1943-1956. In the 1940s and 50s, a handful of scientists from a variety of fields (mathematics, psychology, engineering, economics and political science) began to discuss the possibility of creating an artificial brain. The field of artificial intelligence research was founded as an academic discipline in 1956. The earliest research into thinking machines was inspired by a confluence of ideas that became prevalent in the late 30s, 40s and early 50s. Recent research in neurology had shown that the brain was an electrical network of neurons that fired in all-or-nothing pulses. Norbert Wiener’s cybernetics described control and stability in electrical networks. Claude Shannon’s information theory described digital signals (i.e., all-or-nothing signals). Alan Turing’s theory of computation showed any form computation could be described digitally. The close relationship between these ideas suggested that it might be possible to construct an electronic brain. Examples of work in this vein includes robots such as W. Grey Walter’s turtles and the Johns Hopkins Beast. These machines did not use computers, digital electronics or symbolic reasoning; they were controlled entirely by analog circuitry. Walter Pitts and Warren McCulloch analyzed networks of idealized artificial neurons and showed how they might perform simple logical functions. They were the first to describe what later researchers would call a neural network. One of the students inspired by Pitts and McCulloch was a young Marvin Minsky, then a 24-year old graduate student. In 1951 (with Dean Edmonds) he built the first neural net machine, the SNARC. Minsky was to become one of the most important leaders and innovators in AI for the next 50 years. In 1950 Alan Turing published a landmark paper in which he speculated about the possibility of creating machines that think. He noted that thinking is difficult to define and devised his famous Turing Test. If a machine could carry on a conversation (over a teleprinter) that was indistinguishable from a conversation with a human being, then it was reasonable to say that the machine was thinking. This simplified version of the problem allowed Turing to argue convincingly that a thinking machine was at least plausible and the paper answered all the most common objections to the proposition. The Turing Test was the first serious proposal in the philosophy of artificial intelligence. In 1951, using the Ferranti Mark 1 machine of the University of Manchester, Christopher Strachey wrote a checkers program and Dietrich Prinz wrote one for chess. Arthur Samuel’s checkers program, developed in the middle 50s and early 60s, eventually achieved sufficient skill to challenge a respectable amateur. Game AI would continue to be used as a measure of progress in AI throughout its history.
When access to digital computers became possible in the middle fifties, a few scientists instinctively recognized that a machine that could manipulate numbers could also manipulate symbols and that the manipulation of symbols could well be the essence of human thought. This was a new approach to creating thinking machines. In 1955, Allen Newell and (future Nobel Laureate) Herbert A. Simon created the Logic Theorist (with help from J. C. Shaw). The program would eventually prove 38 of the first 52 theorems in Russell and Whitehead’s Principia Mathematica, and find new and more elegant proofs for some. Simon said that they had solved the venerable mind/body problem, explaining how a system composed of matter can have the properties of mind. (This was an early statement of the philosophical position John Searle would later call Strong AI: that machines can contain minds just as human bodies do.) The Dartmouth Conference of 1956 was organized by Marvin Minsky, John McCarthy and two senior scientists: Claude Shannon and Nathan Rochester of IBM. The proposal for the conference included this assertion: every aspect of learning or any other feature of intelligence can be so precisely described that a machine can be made to simulate it. The participants included Ray Solomonoff, Oliver Selfridge, Trenchard More, Arthur Samuel, Allen Newell and Herbert A. Simon, all of whom would create important programs during the first decades of AI research. At the conference Newell and Simon debuted the Logic Theorist and McCarthy persuaded the attendees to accept Artificial Intelligence as the name of the field. The 1956 Dartmouth conference was the moment that AI gained its name, its mission, its first success and its major players, and is widely considered the birth of AI. The years after the Dartmouth conference were an era of discovery, of sprinting across new ground. The programs that were developed during this time were, to most people, simply astonishing: computers were solving algebra word problems, proving theorems in geometry and learning to speak English. Few at the time would have believed that such intelligent behavior by machines was possible at all. Researchers expressed an intense optimism in private and in print, predicting that a fully intelligent machine would be built in less than 20 years. Government agencies like ARPA poured money into the new field.
There were many successful programs and new directions in the late 50s and 1960s. Many early AI programs used the same basic algorithm. To achieve some goal (like winning a game or proving a theorem), they proceeded step by step towards it (by making a move or a deduction) as if searching through a maze, backtracking whenever they reached a dead end. This paradigm was called reasoning as search. The principal difficulty was that, for many problems, the number of possible paths through the maze was simply astronomical (a situation known as a combinatorial explosion). Researchers would reduce the search space by using heuristics or rules of thumb that would eliminate those paths that were unlikely to lead to a solution. Newell and Simon tried to capture a general version of this algorithm in a program called the General Problem Solver. Other searching programs were able to accomplish impressive tasks like solving problems in geometry and algebra, such as Herbert Gelernter’s Geometry Theorem Prover (1958) and SAINT, written by Minsky’s student James Slagle (1961). Other programs searched through goals and subgoals to plan actions, like the STRIPS system developed at Stanford to control the behavior of their robot Shakey.
An example of a semantic network
An important goal of AI research is to allow computers to communicate in natural languages like English. An early success was Daniel Bobrow’s program STUDENT, which could solve high school algebra word problems. Daniel Bobrow was another brilliant student of Marvin Minsky at MIT. (As it turns out Joyce Hays’ uncle and cousins were/are Bobrows and Jules Mitchel’s uncle and cousins were/are Bobrows. Hopefully, we’ve inherited the Bobrow brilliant genes).
A semantic net represents concepts (e.g. house, door) as nodes and relations among concepts (e.g. has-a) as links between the nodes. The first AI program to use a semantic net was written by Ross Quillian and the most successful (and controversial) version was Roger Schank’s Conceptual dependency theory. Joseph Weizenbaum’s ELIZA could carry out conversations that were so realistic that users occasionally were fooled into thinking they were communicating with a human being and not a program. But in fact, ELIZA had no idea what she was talking about. She simply gave a canned response or repeated back what was said to her, rephrasing her response with a few grammar rules. ELIZA was the first chatterbot. (as in the personal assistant, SIRI, by Apple). In the late 60s, Marvin Minsky and Seymour Papert of the MIT AI Laboratory proposed that AI research should focus on artificially simple situations known as micro-worlds. They pointed out that in successful sciences like physics, basic principles were often best understood using simplified models like frictionless planes or perfectly rigid bodies. Much of the research focused on a blocks world, which consists of colored blocks of various shapes and sizes arrayed on a flat surface. This paradigm led to innovative work in machine vision by Gerald Sussman (who led the team), Adolfo Guzman, David Waltz (who invented constraint propagation), and especially Patrick Winston. At the same time, Minsky and Papert built a robot arm that could stack blocks, bringing the blocks world to life. The crowning achievement of the micro-world program was Terry Winograd’s SHRDLU. It could communicate in ordinary English sentences, plan operations and execute them.
In June 1963, MIT received a $2.2 million grant from the newly created Advanced Research Projects Agency (later known as DARPA). The money was used to fund project MAC which subsumed the AI Group founded by Minsky and McCarthy five years earlier. ARPA continued to provide three million dollars a year until the 70s. ARPA made similar grants to Newell and Simon’s program at CMU and to the Stanford AI Project (founded by John McCarthy in 1963). Another important AI laboratory was established at Edinburgh University by Donald Michie in 1965. These four institutions would continue to be the main centers of AI research (and funding) in academia for many years. The money was proffered with few strings attached: J. C. R. Licklider, then the director of ARPA, believed that his organization should fund people, not projects! and allowed researchers to pursue whatever directions might interest them. This created a freewheeling atmosphere at MIT that gave birth to the hacker culture, but this hands off approach would not last. In the 70s, AI was subject to critiques and financial setbacks. AI researchers had failed to appreciate the difficulty of the problems they faced. Their tremendous optimism had raised expectations impossibly high, and when the promised results failed to materialize, funding for AI disappeared. In the early seventies, the capabilities of AI programs were limited. Even the most impressive could only handle trivial versions of the problems they were supposed to solve; all the programs were, in some sense, toys. AI researchers had begun to run into several fundamental limits that could not be overcome in the 1970s. Although some of these limits would be conquered in later decades, others still stymie the field to this day.
Limited computer power: There was not enough memory or processing speed to accomplish anything truly useful. For example, Ross Quillian’s successful work on natural language was demonstrated with a vocabulary of only twenty words, because that was all that would fit in memory. Hans Moravec argued in 1976 that computers were still millions of times too weak to exhibit intelligence. He suggested an analogy: artificial intelligence requires computer power in the same way that aircraft require horsepower. Below a certain threshold, it’s impossible, but, as power increases, eventually it could become easy. With regard to computer vision, Moravec estimated that simply matching the edge and motion detection capabilities of human retina in real time would require a general-purpose computer capable of 109 operations/second (1000 MIPS). As of 2011, practical computer vision applications require 10,000 to 1,000,000 MIPS. By comparison, the fastest supercomputer in 1976, Cray-1 (retailing at $5 million to $8 million), was only capable of around 80 to 130 MIPS, and a typical desktop computer at the time achieved less than 1 MIPS. Intractability and the combinatorial explosion. In 1972 Richard Karp (building on Stephen Cook’s 1971 theorem) showed there are many problems that can probably only be solved in exponential time (in the size of the inputs). Finding optimal solutions to these problems requires unimaginable amounts of computer time except when the problems are trivial. This almost certainly meant that many of the toy solutions used by AI would probably never scale up into useful systems.
Commonsense knowledge and reasoning: Many important artificial intelligence applications like vision or natural language require simply enormous amounts of information about the world: the program needs to have some idea of what it might be looking at or what it is talking about. This requires that the program know most of the same things about the world that a child does. Researchers soon discovered that this was a truly vast amount of information. No one in 1970 could build a database so large and no one knew how a program might learn so much information.
Moravec’s paradox: Proving theorems and solving geometry problems is comparatively easy for computers, but a supposedly simple task like recognizing a face or crossing a room without bumping into anything is extremely difficult. This helps explain why research into vision and robotics had made so little progress by the middle 1970s.
The frame and qualification problems: AI researchers (like John McCarthy) who used logic discovered that they could not represent ordinary deductions that involved planning or default reasoning without making changes to the structure of logic itself. They developed new logics (like non-monotonic logics and modal logics) to try to solve the problems. The agencies which funded AI research (such as the British government, DARPA and NRC) became frustrated with the lack of progress and eventually cut off almost all funding for undirected research into AI. The pattern began as early as 1966 when the ALPAC report appeared criticizing machine translation efforts. After spending 20 million dollars, the NRC ended all support. In 1973, the Lighthill report on the state of AI research in England criticized the utter failure of AI to achieve its grandiose objectives and led to the dismantling of AI research in that country. (The report specifically mentioned the combinatorial explosion problem as a reason for AI’s failings.) DARPA was deeply disappointed with researchers working on the Speech Understanding Research program at CMU and canceled an annual grant of three million dollars. By 1974, funding for AI projects was hard to find. Hans Moravec blamed the crisis on the unrealistic predictions of his colleagues. Many researchers were caught up in a web of increasing exaggeration. However, there was another issue: since the passage of the Mansfield Amendment in 1969, DARPA had been under increasing pressure to fund mission-oriented direct research, rather than basic undirected research. Funding for the creative, freewheeling exploration that had gone on in the 60s would not come from DARPA. Instead, the money was directed at specific projects with clear objectives, such as autonomous tanks and battle management systems. Several philosophers had strong objections to the claims being made by AI researchers. One of the earliest was John Lucas, who argued that G?del’s incompleteness theorem showed that a formal system (such as a computer program) could never see the truth of certain statements, while a human being could. Hubert Dreyfus ridiculed the broken promises of the 60s and critiqued the assumptions of AI, arguing that human reasoning actually involved very little symbol processing and a great deal of embodied, instinctive, unconscious know how. John Searle’s Chinese Room argument, presented in 1980, attempted to show that a program could not be said to understand the symbols that it uses (a quality called intentionality). If the symbols have no meaning for the machine, Searle argued, then the machine cannot be described as thinking. These critiques were not taken seriously by AI researchers, often because they seemed so far off the point. Problems like intractability and commonsense knowledge seemed much more immediate and serious. It was unclear what difference know how or intentionality made to an actual computer program. Minsky said of Dreyfus and Searle they misunderstand, and should be ignored. Dreyfus, who taught at MIT, was given a cold shoulder: he later said that AI researchers dared not be seen having lunch with me.
Joseph Weizenbaum, the author of ELIZA, felt his colleagues’ treatment of Dreyfus was unprofessional and childish. Although he was an outspoken critic of Dreyfus’ positions, he deliberately made it plain that theirs was not the way to treat a human being. Weizenbaum began to have serious ethical doubts about AI when Kenneth Colby wrote DOCTOR, a chatterbot therapist. Weizenbaum was disturbed that Colby saw his mindless program as a serious therapeutic tool. A feud began, and the situation was not helped when Colby did not credit Weizenbaum for his contribution to the program. In 1976, Weizenbaum published Computer Power and Human Reason which argued that the misuse of artificial intelligence has the potential to devalue human life. A perceptron was a form of neural network introduced in 1958 by Frank Rosenblatt, who had been a schoolmate of Marvin Minsky at the Bronx High School of Science. Like most AI researchers, he was optimistic about their power, predicting that perceptron may eventually be able to learn, make decisions, and translate languages. An active research program into the paradigm was carried out throughout the 60s but came to a sudden halt with the publication of Minsky and Papert’s 1969 book Perceptrons. It suggested that there were severe limitations to what perceptrons could do and that Frank Rosenblatt’s predictions had been grossly exaggerated. The effect of the book was devastating: virtually no research at all was done in connectionism for 10 years. Eventually, a new generation of researchers would revive the field and thereafter it would become a vital and useful part of artificial intelligence. Rosenblatt would not live to see this, as he died in a boating accident shortly after the book was published.
Logic was introduced into AI research as early as 1958, by John McCarthy in his Advice Taker proposal. In 1963, J. Alan Robinson had discovered a simple method to implement deduction on computers, the resolution and unification algorithm. However, straightforward implementations, like those attempted by McCarthy and his students in the late 60s, were especially intractable: the programs required astronomical numbers of steps to prove simple theorems. A more fruitful approach to logic was developed in the 70s by Robert Kowalski at the University of Edinburgh, and soon this led to the collaboration with French researchers Alain Colmerauer and Philippe Roussel, who created the successful logic programming language Prolog. Prolog uses a subset of logic (Horn clauses, closely related to rules and production rules) that permit tractable computation. Rules would continue to be influential, providing a foundation for Edward Feigenbaum’s expert systems and the continuing work by Allen Newell and Herbert A. Simon that would lead to Soar and their unified theories of cognition. Critics of the logical approach noted, as Dreyfus had, that human beings rarely used logic when they solved problems. Experiments by psychologists like Peter Wason, Eleanor Rosch, Amos Tversky, Daniel Kahneman and others provided proof. McCarthy responded that what people do is irrelevant. He argued that what is really needed are machines that can solve problems – not machines that think as people do. Among the critics of McCarthy’s approach were his colleagues across the country at MIT. Marvin Minsky, Seymour Papert and Roger Schank were trying to solve problems like story understanding and object recognition that required a machine to think like a person. In order to use ordinary concepts like chair or restaurant they had to make all the same illogical assumptions that people normally made. Unfortunately, imprecise concepts like these are hard to represent in logic. Gerald Sussman observed that using precise language to describe essentially imprecise concepts doesn’t make them any more precise. In 1975, in a seminal paper, Minsky noted that many of his fellow scruffy researchers were using the same kind of tool: a framework that captures all our common sense assumptions about something. For example, if we use the concept of a bird, there is a constellation of facts that immediately come to mind: we might assume that it flies, eats worms and so on. We know these facts are not always true and that deductions using these facts will not be logical, but these structured sets of assumptions are part of the context of everything we say and think. He called these structures frames. Schank used a version of frames he called scripts to successfully answer questions about short stories in English. Many years later object-oriented programming would adopt the essential idea of inheritance from AI research on frames. Today, our everyday language has absorbed certain computer terms, like frame. For example we talk about, framing an argument or framing a point of view. In the 1980s a form of AI program called expert systems was adopted by corporations around the world and knowledge became the focus of mainstream AI research. In those same years, the Japanese government aggressively funded AI with its fifth generation computer project. Another encouraging event in the early 1980s was the revival of connectionism in the work of John Hopfield and David Rumelhart. Once again, AI had achieved success.
An expert system is a program that answers questions or solves problems about a specific domain of knowledge, using logical rules that are derived from the knowledge of experts. The earliest examples were developed by Edward Feigenbaum and his students. Dendral, begun in 1965, identified compounds from spectrometer readings. MYCIN, developed in 1972, diagnosed infectious blood diseases. They demonstrated the feasibility of the approach. Expert systems restricted themselves to a small domain of specific knowledge (thus avoiding the commonsense knowledge problem) and their simple design made it relatively easy for programs to be built and then modified once they were in place. All in all, the programs proved to be useful: something that AI had not been able to achieve up to this point. In 1980, an expert system called XCON was completed at CMU for the Digital Equipment Corporation. It was an enormous success: it was saving the company 40 million dollars annually by 1986. Corporations around the world began to develop and deploy expert systems and by 1985 they were spending over a billion dollars on AI, most of it to in-house AI departments. An industry grew up to support them, including hardware companies like Symbolics and Lisp Machines and software companies such as IntelliCorp and Aion. The power of expert systems came from the expert knowledge they contained. They were part of a new direction in AI research that had been gaining ground throughout the 70s. AI researchers were beginning to suspect – reluctantly, for it violated the scientific canon of parsimony – that intelligence might very well be based on the ability to use large amounts of diverse knowledge in different ways. The great lesson from the 1970s was that intelligent behavior depended very much on dealing with knowledge, sometimes quite detailed knowledge, of a domain where a given task lay. Knowledge based systems and knowledge engineering became a major focus of AI research in the 1980s. The 1980s also saw the birth of Cyc, the first attempt to attack the commonsense knowledge problem directly, by creating a massive database that would contain all the mundane facts that the average person knows. Douglas Lenat, who started and led the project, argued that there is no shortcut – the only way for machines to know the meaning of human concepts is to teach them, one concept at a time, by hand. The project was not expected to be completed for many decades.
In 1981, the Japanese Ministry of International Trade and Industry set aside $850 million for the Fifth generation computer project. Their objectives were to write programs and build machines that could carry on conversations, translate languages, interpret pictures, and reason like human beings. Much to the chagrin of scruffies, they chose Prolog as the primary computer language for the project. Other countries responded with new programs of their own. The UK began the Alvey project. A consortium of American companies formed the Microelectronics and Computer Technology Corporation (or MCC) to fund large scale projects in AI and information technology. DARPA responded as well, founding the Strategic Computing Initiative and tripling its investment in AI between 1984 and 1988.
A Hopfield net with four nodes.
In 1982, physicist John Hopfield was able to prove that a form of neural network (now called a Hopfield net) could learn and process information in a completely new way. Around the same time, David Rumelhart popularized a new method for training neural networks called backpropagation (discovered years earlier by Paul Werbos). These two discoveries revived the field of connectionism which had been largely abandoned since 1970. The new field was unified and inspired by the appearance of Parallel Distributed Processing in 1986 – a two volume collection of papers edited by Rumelhart and psychologist James McClelland. Neural networks would become commercially successful in the 1990s, when they began to be used as the engines driving programs like optical character recognition and speech recognition. The business community’s fascination with AI rose and fell in the 80s in the classic pattern of an economic bubble. The collapse was in the perception of AI by government agencies and investors – the field continued to make advances despite the criticism. Rodney Brooks and Hans Moravec, researchers from the related field of robotics, argued for an entirely new approach to artificial intelligence. The term AI winter was coined by researchers who had survived the funding cuts of 1974 when they became concerned that enthusiasm for expert systems had spiraled out of control and that disappointment would certainly follow. Their fears were well founded: in the late 80s and early 90s, AI suffered a series of financial setbacks. The first indication of a change in weather was the sudden collapse of the market for specialized AI hardware in 1987. Desktop computers from Apple, Microsoft and IBM had been steadily gaining speed and power and in 1987 they became more powerful than the more expensive Lisp machines made by Symbolics and others. There was no longer a good reason to buy them. An entire industry worth half a billion dollars was demolished overnight. Eventually the earliest successful expert systems, such as XCON, proved too expensive to maintain. They were difficult to update, they could not learn, they were brittle (i.e., they could make grotesque mistakes when given unusual inputs), and they fell prey to problems (such as the qualification problem) that had been identified years earlier. Expert systems proved useful, but only in a few special contexts.
In the late 80s, the Strategic Computing Initiative cut funding to AI deeply and brutally. New leadership at DARPA had decided that AI was not the next wave and directed funds towards projects that seemed more likely to produce immediate results. By 1991, the impressive list of goals penned in 1981 for Japan’s Fifth Generation Project had not been met. Indeed, some of them, like carry on a casual conversation had not been met by 2010. As with other AI projects, expectations had run much higher than what was actually possible. In the late 80s, several researchers advocated a completely new approach to artificial intelligence, based on robotics. They believed that, to show real intelligence, a machine needs to have a body – it needs to perceive, move, survive and deal with the world. They argued that these sensorimotor skills are essential to higher level skills like commonsense reasoning and that abstract reasoning was actually the least interesting or important human skill (see Moravec’s paradox). They advocated building intelligence from the bottom up. The approach revived ideas from cybernetics and control theory that had been unpopular since the sixties. Another precursor was David Marr, who had come to MIT in the late 70s from a successful background in theoretical neuroscience to lead the group studying vision. He rejected all symbolic approaches (both McCarthy’s logic and Minsky’s frames), arguing that AI needed to understand the physical machinery of vision from the bottom up before any symbolic processing took place. (Marr’s work would be cut short by leukemia in 1980.)
In a 1990 paper Elephants Don’t Play Chess, robotics researcher Rodney Brooks took direct aim at the physical symbol system hypothesis, arguing that symbols are not always necessary since the world is its own best model. It is always exactly up to date. It always has every detail there is to be known. The trick is to sense it appropriately and often enough. In the 80s and 90s, many cognitive scientists also rejected the symbol processing model of the mind and argued that the body was essential for reasoning, a theory called the embodied mind thesis. The field of AI, now more than a half a century old, finally achieved some of its oldest goals. It began to be used successfully throughout the technology industry, although somewhat behind the scenes. Some of the success was due to increasing computer power and some was achieved by focusing on specific isolated problems and pursuing them with the highest standards of scientific accountability. Still, the reputation of AI, in the business world at least, was less than pristine. Inside the field there was little agreement on the reasons for AI’s failure to fulfill the dream of human level intelligence that had captured the imagination of the world in the 1960s. Together, all these factors helped to fragment AI into competing subfields focused on particular problems or approaches, sometimes even under new names that disguised the tarnished pedigree of artificial intelligence.AI was both more cautious and more successful than it had ever been. On 11 May 1997, Deep Blue became the first computer chess-playing system to beat a reigning world chess champion, Garry Kasparov. In February 2011, in a Jeopardy! quiz show exhibition match, IBM’s question answering system, Watson, defeated the two greatest Jeopardy! champions, Brad Rutter and Ken Jennings, by a significant margin. These successes were not due to some revolutionary new paradigm, but mostly on the tedious application of engineering skill and on the tremendous power of computers today. In fact, Deep Blue’s computer was 10 million times faster than the Ferranti Mark 1 that Christopher Strachey taught to play chess in 1951. This dramatic increase is measured by Moore’s law, which predicts that the speed and memory capacity of computers doubles every two years. The fundamental problem of raw computer power was slowly being overcome.
A new paradigm called intelligent agents became widely accepted during the 90s. Although earlier researchers had proposed modular divide and conquer approaches to AI, the intelligent agent did not reach its modern form until Judea Pearl, Allen Newell and others brought concepts from decision theory and economics into the study of AI. When the economist’s definition of a rational agent was married to computer science’s definition of an object or module, the intelligent agent paradigm was complete. An intelligent agent is a system that perceives its environment and takes actions which maximize its chances of success. By this definition, simple programs that solve specific problems are intelligent agents, as are human beings and organizations of human beings, such as firms. The intelligent agent paradigm defines AI research as the study of intelligent agents. This is a generalization of some earlier definitions of AI: it goes beyond studying human intelligence; it studies all kinds of intelligence. The paradigm gave researchers license to study isolated problems and find solutions that were both verifiable and useful. It provided a common language to describe problems and share their solutions with each other, and with other fields that also used concepts of abstract agents, like economics and control theory. It was hoped that a complete agent architecture (like Newell’s SOAR) would one day allow researchers to build more versatile and intelligent systems out of interacting intelligent agents. AI researchers began to develop and use sophisticated mathematical tools more than they ever had in the past. There was a widespread realization that many of the problems that AI needed to solve were already being worked on by researchers in fields like mathematics, economics or operations research. The shared mathematical language allowed both a higher level of collaboration with more established and successful fields and the achievement of results which were measurable and provable; AI had become a more rigorous scientific discipline. Russell & Norvig (2003) describe this as nothing less than a revolution and the victory of the neats. Judea Pearl’s highly influential 1988 book brought probability and decision theory into AI. Among the many new tools in use were Bayesian networks, hidden Markov models, information theory, stochastic modeling and classical optimization. Precise mathematical descriptions were also developed for computational intelligence paradigms like neural networks and evolutionary algorithms. Algorithms originally developed by AI researchers began to appear as parts of larger systems. AI had solved a lot of very difficult problems and their solutions proved to be useful throughout the technology industry, such as data mining, industrial robotics, logistics, speech recognition, banking software, medical diagnosis and Google’s search engine.
The field of AI receives little or no credit for these successes. Many of AI’s greatest innovations have been reduced to the status of just another item in the tool chest of computer science. Nick Bostrom explains A lot of cutting edge AI has filtered into general applications, often without being called AI because once something becomes useful enough and common enough it’s not labeled AI anymore. Many researchers in AI in 1990s deliberately called their work by other names, such as informatics, knowledge-based systems, cognitive systems or computational intelligence. In part, this may be because they considered their field to be fundamentally different from AI, but also the new names help to procure funding. In the commercial world at least, the failed promises of the AI Winter continue to haunt AI research, as the New York Times reported in 2005: Computer scientists and software engineers avoided the term artificial intelligence for fear of being viewed as wild-eyed dreamers. Where is HAL 9000? In 1968, Arthur C. Clarke and Stanley Kubrick had imagined that by the year 2001, a machine would exist with an intelligence that matched or exceeded the capability of human beings. The character they created, HAL 9000, was based on a belief shared by many leading AI researchers that such a machine would exist by the year 2001. Marvin Minsky asks So the question is why didn’t we get HAL in 2001? Minsky believes that the answer is that the central problems, like commonsense reasoning, were being neglected, while most researchers pursued things like commercial applications of neural nets or genetic algorithms. John McCarthy, on the other hand, still blames the qualification problem. For Ray Kurzweil, the issue is computer power and, using Moore’s Law, he predicts that machines with human-level intelligence will appear by 2029. Jeff Hawkins argues that neural net research ignores the essential properties of the human cortex, preferring simple models that have been successful at solving simple problems. There are many other explanations and for each there is a corresponding research program underway. In February 2011, in a Jeopardy! quiz show exhibition match, IBM’s question answering system, Watson, defeated the two greatest Jeopardy champions, Brad Rutter and Ken Jennings, by a significant margin. The Kinect, which provides a 3D body-motion interface for the Xbox 360 and the Xbox One, uses algorithms that emerged from lengthy AI research as do intelligent personal assistants in smartphones. In 2015 researchers constructed a robot, which learned how to use some tools by watching YouTube videos. After winning Jeopardy, Watson wasted no time. He proceeded to Memorial Sloan Kettering Hospital in Manhattan for training as a physician’s assistant. After completing this sophisticated level of AI in
Plasma 25-hydroxyvitamin D and Colorectal Cancer Risk
A study published online in the journal GUT (16 January 2015) has hypothesized that the inverse association between vitamin D and colorectal cancer risk might be stronger for cancers with high-level immune response than those with low-level immune response.
Evidence has suggested the protective effects of vitamin D and antitumor immunity on colorectal cancer risk. The mechanism seems to be that the immune cells in tumor microenvironment, can convert 25-hydroxyvitamin D [25(OH)D] to bioactive 1a,25-dihydroxyvitamin D3, which influences neoplastic and immune cells as an autocrine and paracrine factor. Autocrine signaling is a form of cell signaling in which a cell secretes a hormone or chemical messenger (called the autocrine agent) that binds to autocrine receptors on that same cell, leading to changes in the cell. An example of an autocrine agent is when the cytokine interleukin-1 (IL-1) is produced in monocytes in response to external stimuli, and IL-1 then binds to cell-surface receptors on the same cell that produced it. This can be contrasted with paracrine signaling where the signaling affects cells of a different type than the cell performing the secretion.
The investigation was a nested case-control study (318 rectal and colon carcinoma cases and 624 matched controls) within the Nurses’ Health Study and Health Professionals Follow-up Study using molecular pathological epidemiology database. Multivariable conditional logistic regression was used to assess the association of plasma 25(OH)D with tumor subtypes according to the degree of lymphocytic reaction, tumor-infiltrating T cells (CD3+, CD8+, CD45RO+ (PTPRC) and FOXP3+ cells), microsatellite instability or CpG island methylator phenotype.
Results showed an association of plasma 25(OH)D with colorectal carcinoma differed by the degree of intratumoral periglandular reaction (p for heterogeneity=0.001); high 25(OH)D was associated with lower tumor risk with high-level reaction (comparing the highest versus lowest tertile: p for trend<0.001), but not risk of tumor with lower-level reaction. A statistically non-significant difference was observed for the associations of 25(OH)D with tumor subtypes according to CD3+ T cell density
The authors concluded that high plasma 25(OH)D level is associated with lower risk of colorectal cancer with intense immune reaction, supporting a role of vitamin D in cancer immunoprevention through tumor-host interaction.
Risk for Alcohol-Medication Interactions
Based on recent estimates, about 71% of U.S. adults drink alcohol.
There is general agreement that older adults are at particular risk of experiencing alcohol-medication interactions. Not only are they more likely to be taking medications in general, but certain alcohol-interactive medications, such as diazepam (Valium), are metabolized more slowly as one ages, creating a larger window for potential interactions.
According to a paper just published online in Alcoholism: Clinical and Experimental Research (16 January 2015), based on a study from the NIH, nearly 42% of U.S. adults who drink also report using medications known to interact with alcohol. Among those over 65 years of age who drink alcohol, nearly 78% report using alcohol-interactive medications. Such medications are widely used, prescribed for common conditions such as depression, diabetes and high blood pressure. The research is among the first to estimate the proportion of adult drinkers in the United States who may be mixing alcohol-interactive medications with alcohol. The main types of alcohol-interactive medications reported in the survey were blood pressure medications, sleeping pills, pain medications, muscle relaxers, diabetes and cholesterol medications, antidepressants and antipsychotics; with resulting health effects ranging from mild (nausea, headaches, loss of coordination) to severe (internal bleeding, heart problems, difficulty breathing).
For the study, the authors analyzed data from more than 26,000 adults ages 20 and older who participated in the National Health and Nutrition Examination Survey (1999-2010). The survey asked participants about alcohol use in the past year and prescription drug use in the past month. The authors noted that the results of the study indicate potential (rather than actual) rates, because it was not possible to confirm whether drinking and medication use overlapped based on the available data. However, it is likely that those who drink regularly and take medication regularly are doing so in a similar time frame.
According to the authors, the findings show that a substantial percentage of those who drink regularly, particularly older adults, could be at risk of harmful alcohol and medication interactions. As a result, it is recommended that people talk to their doctor or pharmacist about whether they should avoid alcohol while taking their prescribed medications.”
For more on alcohol-medication interactions, see the NIAAA fact sheet.
Australia, Brazil, Canada, Japan, and the US: Safeguarding Medical Devices
The FDA and its regulatory counterparts abroad have the responsibility of ensuring the safety of the many regulated medical devices imported in their countries each year. To make this task more manageable, FDA and regulatory agencies in Australia, Brazil, Canada, and Japan embarked in 2014 on a pilot called the Medical Device Single Audit Program (MDSAP), with the goal to develop processes that allows a single audit, or inspection to ensure the medical device regulatory requirements for all five countries are satisfied, in an efficient yet thorough manner.
On January 1, 2015 the MDSAP pilot reached a major milestone – manufacturers around the globe interested in marketing medical devices in Australia, Brazil, Canada, and the U.S. were invited to participate in the program. This summer, when Japan enters the MDSAP as a full member, the same invitation will be issued also to medical device manufacturers interested in marketing in Japan. Under this pilot, audits will be conducted by recognized third-party organizations, and medical device regulators in the participating countries will be able to use these inspection reports when making their regulatory decisions. Not only does this program reduce the participating regulators’ need to individually perform routine inspections; it allows them all to have the same reliable information about inspectional findings. Manufacturers, too, can benefit from the MDSAP pilot by cutting down on the number of regulatory audits they have to host, thereby minimizing manufacturing plant and personnel disruptions. This form of international and standardized oversight lessens the burden on manufacturers by bringing more consistency and transparency to the regulatory process.
The MDSAP pilot does not increase regulatory requirements for medical device manufacturers – the audits cover only the existing requirements of the regulatory authorities participating. In many cases, these requirements are already harmonized or very similar to one another, such as the international standard for medical devices quality management systems (ISO 13485:2003), the Brazilian Good Manufacturing Practices (RDC ANVISA 16/2013), the U.S. Quality System Regulation (21 CFR Part 820), and other specific pre- and post-market regulatory requirements of the authorities participating in the MDSAP pilot.
The FDA will accept MDSAP audits as a substitute for routine FDA inspections, typically done every two years for all classes of medical devices and including in vitro diagnostic devices. Pre-approval inspections for devices requiring premarket approval applications (PMAs) and for cause compliance inspections will not be part of the MDSAP pilot. Manufacturers that choose to participate in the pilot program will help to shape the policies and procedures of the fully operational MDSAP, which is scheduled to begin in 2017. We expect that the MDSAP pilot will enhance confidence in third party audit programs, increasing the footprint of this global endeavor.
According to FDA, international cooperation promotes global alignment of regulatory approaches and technical requirements, expanding the safety net that protects patients world-wide.
New information about how countries will participate in the MDSAP pilot is available on the FDA’s MDSAP pilot web page. Manufacturers can find additional information on the MDSAP web page. This MDSAP page provides information on the auditing organizations involved in the pilot for interested manufacturers to contact directly.