What is Artificial Intelligence?
Now our Artificial Intelligence — trainer Norbu is being self-taught by collected data and AE-trainer will create individual programs.
Can machines think? — Alan Turing, 1950
Less than a decade after breaking the Nazi encryption machine Enigma and helping the Allied Forces win World War II, mathematician Alan Turing changed history a second time with a simple question: “Can machines think?”
Turing’s paper “Computing Machinery and Intelligence” (1950), and it’s subsequent Turing Test, established the fundamental goal and vision of artificial intelligence.
At it’s core, AI is the branch of computer science that aims to answer Turing’s question in the affirmative. It is the endeavor to replicate or simulate human intelligence in machines.
The expansive goal of artificial intelligence has given rise to many questions and debates. So much so, that no singular definition of the field is universally accepted.
The major limitation in defining AI as simply “building machines that are intelligent” is that it doesn’t actually explain what artificial intelligence is? What makes a machine intelligent?
In their groundbreaking textbook Artificial Intelligence: A Modern Approach, authors Stuart Russell and Peter Norvig approach the question by unifying their work around the theme of intelligent agents in machines. With this in mind, AI is “the study of agents that receive percepts from the environment and perform actions.” (Russel and Norvig viii)
Norvig and Russell go on to explore four different approaches that have historically defined the field of AI:
- Thinking humanly
- Thinking rationally
- Acting humanly
- Acting rationally
The first two ideas concern thought processes and reasoning, while the others deal with behavior. Norvig and Russell focus particularly on rational agents that act to achieve the best outcome, noting “all the skills needed for the Turing Test also allow an agent to act rationally.” (Russel and Norvig 4).
Patrick Winston, the Ford professor of artificial intelligence and computer science at MIT, defines AI as “algorithms enabled by constraints, exposed by representations that support models targeted at loops that tie thinking, perception and action together.”
While addressing a crowd at the Japan AI Experience in 2017, DataRobot CEO Jeremy Achin began his speech by offering the following definition of how AI is used today:
“AI is a computer system able to perform tasks that ordinarily require human intelligence… Many of these artificial intelligence systems are powered by machine learning, some of them are powered by deep learning and some of them are powered by very boring things like rules.”
Artificial intelligence generally falls under two broad categories:
- Narrow AI: Sometimes referred to as “Weak AI,” this kind of artificial intelligence operates within a limited context and is a simulation of human intelligence. Narrow AI is often focused on performing a single task extremely well and while these machines may seem intelligent, they are operating under far more constraints and limitations than even the most basic human intelligence.
- Artificial General Intelligence (AGI): AGI, sometimes referred to as “Strong AI,” is the kind of artificial intelligence we see in the movies, like the robots from Westworld or Data from Star Trek: The Next Generation. AGI is a machine with general intelligence and, much like a human being, it can apply that intelligence to solve any problem.
ARTIFICIAL INTELLIGENCE EXAMPLES
- Smart assistants (like Siri and Alexa)
- Disease mapping and prediction tools
- Manufacturing and drone robots
- Optimized, personalized healthcare treatment recommendations
- Conversational bots for marketing and customer service
- Robo-advisors for stock trading
- Spam filters on email
- Social media monitoring tools for dangerous content or false news
- Song or TV show recommendations from Spotify and Netflix
Narrow Artificial Intelligence
Narrow AI is all around us and is easily the most successful realization of artificial intelligence to date. With its focus on performing specific tasks, Narrow AI has experienced numerous breakthroughs in the last decade that have had “significant societal benefits and have contributed to the economic vitality of the nation,” according to “Preparing for the Future of Artificial Intelligence,” a 2016 report released by the Obama Administration.
A few examples of Narrow AI include:
- Google search
- Image recognition software
- Siri, Alexa and other personal assistants
- Self-driving cars
- IBM’s Watson
Machine Learning & Deep Learning
Much of Narrow AI is powered by breakthroughs in machine learning and deep learning. Understanding the difference between artificial intelligence, machine learning and deep learning can be confusing. Venture capitalist Frank Chen provides a good overview of how to distinguish between them, noting:
“Artificial intelligence is a set of algorithms and intelligence to try to mimic human intelligence. Machine learning is one of them, and deep learning is one of those machine learning techniques.”
Simply put, machine learning feeds a computer data and uses statistical techniques to help it “learn” how to get progressively better at a task, without having been specifically programmed for that task, eliminating the need for millions of lines of written code. Machine learning consists of both supervised learning (using labeled data sets) and unsupervised learning (using unlabeled data sets).
Deep learning is a type of machine learning that runs inputs through a biologically-inspired neural network architecture. The neural networks contain a number of hidden layers through which the data is processed, allowing the machine to go “deep” in its learning, making connections and weighting input for the best results.
Artificial General Intelligence
The creation of a machine with human-level intelligence that can be applied to any task is the Holy Grail for many AI researchers, but the quest for AGI has been fraught with difficulty.
The search for a “universal algorithm for learning and acting in any environment,” (Russel and Norvig 27) isn’t new, but time hasn’t eased the difficulty of essentially creating a machine with a full set of cognitive abilities.
AGI has long been the muse of dystopian science fiction, in which super-intelligent robots overrun humanity, but experts agree it’s not something we need to worry about anytime soon.
Intelligent robots and artificial beings first appeared in the ancient Greek myths of Antiquity. Aristotle’s development of the syllogism and it’s use of deductive reasoning was a key moment in mankind’s quest to understand its own intelligence. While the roots are long and deep, the history of artificial intelligence as we think of it today spans less than a century. The following is a quick look at some of the most important events in AI.
- Warren McCullough and Walter Pitts publish “A Logical Calculus of Ideas Immanent in Nervous Activity.” The paper proposed the first mathematic model for building a neural network.
- In his book The Organization of Behavior: A Neuropsychological Theory, Donald Hebb proposes the theory that neural pathways are created from experiences and that connections between neurons become stronger the more frequently they’re used. Hebbian learning continues to be an important model in AI.
- Alan Turing publishes “Computing Machinery and Intelligence, proposing what is now known as the Turing Test, a method for determining if a machine is intelligent.
- Harvard undergraduates Marvin Minsky and Dean Edmonds build SNARC, the first neural network computer.
- Claude Shannon publishes the paper “Programming a Computer for Playing Chess.”
- Isaac Asimov publishes the “Three Laws of Robotics.”
- Arthur Samuel develops a self-learning program to play checkers.
- The Georgetown-IBM machine translation experiment automatically translates 60 carefully selected Russian sentences into English.
- The phrase artificial intelligence is coined at the “Dartmouth Summer Research Project on Artificial Intelligence.” Led by John McCarthy, the conference, which defined the scope and goals of AI, is widely considered to be the birth of artificial intelligence as we know it today.
- Allen Newell and Herbert Simon demonstrate Logic Theorist (LT), the first reasoning program.
- John McCarthy develops the AI programming language Lisp and publishes the paper “Programs with Common Sense.” The paper proposed the hypothetical Advice Taker, a complete AI system with the ability to learn from experience as effectively as humans do.
- Allen Newell, Herbert Simon and J.C. Shaw develop the General Problem Solver (GPS), a program designed to imitate human problem-solving.
- Herbert Gelernter develops the Geometry Theorem Prover program.
- Arthur Samuel coins the term machine learning while at IBM.
- John McCarthy and Marvin Minsky found the MIT Artificial Intelligence Project.
- John McCarthy starts the AI Lab at Stanford.
- The Automatic Language Processing Advisory Committee (ALPAC) report by the U.S. government details the lack of progress in machine translations research, a major Cold War initiative with the promise of automatic and instantaneous translation of Russian. The ALPAC report leads to the cancellation of all government-funded MT projects.
- The first successful expert systems are developed in DENDRAL, a XX program, and MYCIN, designed to diagnose blood infections, are created at Stanford.
- The logic programming language PROLOG is created.
- The “Lighthill Report,” detailing the disappointments in AI research, is released by the British government and leads to severe cuts in funding for artificial intelligence projects.
- Frustration with the progress of AI development leads to major DARPA cutbacks in academic grants. Combined with the earlier ALPAC report and the previous year’s “Lighthill Report,” artificial intelligence funding dries up and research stalls. This period is known as the “First AI Winter.”
- Digital Equipment Corporations develops R1 (also known as XCON), the first successful commercial expert system. Designed to configure orders for new computer systems, R1 kicks off an investment boom in expert systems that will last for much of the decade, effectively ending the first “AI Winter.”
- Japan’s Ministry of International Trade and Industry launches the ambitious Fifth Generation Computer Systems project. The goal of FGCS is to develop supercomputer-like performance and a platform for AI development.
- In response to Japan’s FGCS, the U.S. government launches the Strategic Computing Initiative to provide DARPA funded research in advanced computing and artificial intelligence.
- Companies are spending more than a billion dollars a year on expert systems and an entire industry known as the Lisp machine market springs up to support them. Companies like Symbolics and Lisp Machines Inc. build specialized computers to run on the AI programming language Lisp.
- As computing technology improved, cheaper alternatives emerged and the Lisp machine market collapsed in 1987, ushering in the “Second AI Winter.” During this period, expert systems proved too expensive to maintain and update, eventually falling out of favor.
- Japan terminates the FGCS project in 1992, citing failure in meeting the ambitious goals outlined a decade earlier.
- DARPA ends the Strategic Computing Initiative in 1993 after spending nearly $1 billion and falling far short of expectations.
- U.S. forces deploy DART, an automated logistics planning and scheduling tool, during the Gulf War.
- IBM’s Deep Blue beats world chess champion Gary Kasparov
- STANLEY, a self-driving car, wins the DARPA Grand Challenge.
- The U.S. military begins investing in autonomous robots like Boston Dynamic’s “Big Dog” and iRobot’s “PackBot.”
- Google makes breakthroughs in speech recognition and introduces the feature in its iPhone app.
- IBM’s Watson trounces the competition on Jeopardy!.
- Andrew Ng, founder of the Google Brain Deep Learning project, feeds a neural network using deep learning algorithms 10 million YouTube videos as a training set. The neural network learned to recognize a cat without being told what a cat is, ushering in breakthrough era for neural networks and deep learning funding.
- Google makes first self-driving car to pass a state driving test.
- Google DeepMind’s AlphaGo defeats world champion Go player Lee Sedol. The complexity of the ancient Chinese game was seen as a major hurdle to clear in AI.
Stress isn’t always a bad thing
It can give us extra energy for achievements, but if we face stress everyday it actually begins to change our brain. Find out more in this video
How changing what you eat could reduce your stress
What you eat could improve how your body responds to stress, according to experts. Introducing some new foods while cutting back on others may reduce the impact stress has on your body and day-to-day life. Here are four dietary changes designed to help you handle stress better.
As well as feeding you, the food you eat feeds the trillions of bacteria that live in your gut. Some studies find regulating gut bacteria via diet can have a positive impact on anxiety symptoms. The amount and types of bacteria in your gut are affected by your diet, and good dietary choices can “communicate calm to the brain”, according to Dr Rangan Chaterjee.
“The cheapest and simplest thing you can do to diversify and optimise the ‘good’ bacteria in your gut is to increase the variety of foods you eat…. Try to eat five different coloured vegetables every day”, says Dr Chaterjee. That’s because “minimally processed wholefoods – fibre” are best for your gut bacteria.
How does it work? The body cannot digest some fibre, including that found in many fruits and vegetables, so it’s fermented in the gut. Here it supports the growth of helpful microbes. As part of the fermentation process, the bacteria release essential chemicals and acids that interact with all the cells in your body. This interaction influences brain and immune health, according to Felice Jacka, Head of the Food and Mood Centre at Deakin University in Australia.
Jacka recommends eating foods from the Mediterranean diet, such as “fruit, vegetables, legumes and wholegrains, which all contain plant fibre”. She says fermented foods such as sauerkraut, kimchi, kombucha and kefir can nurture bacteria and yeasts in the gut, as well as being good for your wider health.
Step away from the sugar
If your idea of dealing with stress is downing a pint of ice cream, you’re probably familiar with the concept of a ‘sugar high’, and possibly the ‘crash’ that comes afterwards. When you eat lots of sugar, your body releases insulin to absorb the excess glucose and regulate your blood-sugar levels. This journey of ups and downs may have negative side-effects, such as lack of concentration and tiredness, which can “hinder your ability to cope with stressful situations”, according to dietitian Sophie Medlin.
Work out when you’ve had enough caffeine
Positive and negative effects of caffeine on anxiety have been recorded. Most studies are based on research done with people already diagnosed with anxiety, but dietician Medlin says the results can often be applied to people suffering from everyday stress.
“The way you react may depend on your genetic sensitivity to caffeine, your gender and how much you drink, as well as existing anxious feelings”, says Medlin. Drinking caffeine can affect your sleep, which may impact the way you deal with stress. If drinking or eating caffeine negatively affects you, it’s a good idea to switch to decaf (although do this slowly, as caffeine withdrawals are no joke!).
Don’t mix hunger and stress
“Never go into any stressful situation hungry… we’ve long recognised that hunger leads to irritability”, says Medlin. When you’re hungry, your blood sugar drops and your cortisol and adrenaline – those fight-or-flight hormones – rise. This is because the neuropeptides, secreted by neurons to control the chemicals in the brain, are the same for hunger as for anger. “Your body is pretty much still a caveman”, says Medlin. This explains why we often become irritable when peckish.
Adding ‘hanger’ to a stressful situation can enhance negative emotions. Have a few healthy snacks at hand to keep hunger pangs at bay, especially if you’re expecting a trying day.
Researchers study how it seems to change the brain in depressed patients
“In 2015, 16.1 million Americans reported experiencing major depression during the previous year, often struggling to function while grappling with crippling darkness and despair.”
There’s an arsenal of treatments at hand, including talk therapy and antidepressant medications, but what’s depressing in itself is that they don’t work for every patient.
“Many people don’t respond to the frontline interventions,” said Benjamin Shapero, an instructor in psychiatry at Harvard Medical School (HMS) and a psychologist at Massachusetts General Hospital’s (MGH) Depression Clinical and Research Program. “Individual cognitive behavioral therapy is helpful for many people; antidepressant medications help many people. But it’s also the case that many people don’t benefit from them as well. There’s a great need for alternative approaches.”
Shapero is working with Gaëlle Desbordes, an instructor in radiology at HMS and a neuroscientist at MGH’s Martinos Center for Biomedical Imaging, to explore one alternative approach: mindfulness-based meditation.
In recent decades, public interest in mindfulness meditation has soared. Paralleling, and perhaps feeding, the growing popular acceptance has been rising scientific attention. The number of randomized controlled trials — the gold standard for clinical study — involving mindfulness has jumped from one in the period from 1995‒1997 to 11 from 2004‒2006, to a whopping 216 from 2013‒2015, according to a recent article summarizing scientific findings on the subject.
Studies have shown benefits against an array of conditions both physical and mental, including irritable bowel syndrome, fibromyalgia, psoriasis, anxiety, depression, and post-traumatic stress disorder. But some of those findings have been called into question because studies had small sample sizes or problematic experimental designs. Still, there are a handful of key areas — including depression, chronic pain, and anxiety — in which well-designed, well-run studies have shown benefits for patients engaging in a mindfulness meditation program, with effects similar to other existing treatments.
“There are a few applications where the evidence is believable. But the effects are by no means earth-shattering,” Desbordes said. “We’re talking about moderate effect size, on par with other treatments, not better. And then there’s a bunch of other things under study with preliminary evidence that is encouraging but by no means conclusive. I think that’s where it’s at. I’m not sure that is exactly how the public understands it at this point.”
Desbordes’ interest in the topic stems from personal experience. She began meditating as a graduate student in computational neuroscience at Boston University, seeking respite from the stress and frustration of academic life. Her experience convinced her that something real was happening to her and prompted her to study the subject more closely, in hopes of shedding enough light to underpin therapy that might help others.
“My own interest comes from having practiced those [meditation techniques] and found them beneficial, personally. Then, being a scientist, asking ‘How does this work? What is this doing to me?’ and wanting to understand the mechanisms to see if it can help others,” Desbordes said. “If we want that to become a therapy or something offered in the community, we need to demonstrate [its benefits] scientifically.”
Desbordes’ research uses functional magnetic resonance imaging (fMRI), which not only takes pictures of the brain, as a regular MRI does, but also records brain activity occurring during the scan. In 2012, she demonstrated that changes in brain activity in subjects who have learned to meditate hold steady even when they’re not meditating. Desbordes took before-and-after scans of subjects who learned to meditate over the course of two months. She scanned them not while they were meditating, but while they were performing everyday tasks. The scans still detected changes in the subjects’ brain activation patterns from the beginning to the end of the study, the first time such a change — in a part of the brain called the amygdala — had been detected.
In her current work, she is exploring meditation’s effects on the brains of clinically depressed patients, a group for whom studies have shown meditation to be effective. Working with patients selected and screened by Shapero, Desbordes is performing functional magnetic resonance imaging scans before and after an eight-week course in mindfulness-based cognitive therapy, or MBCT.
During the scans, participants complete two tests, one that encourages them to become more aware of their bodies by focusing on their heartbeats (an exercise related to mindfulness meditation), and the other asking them to reflect on phrases common in the self-chatter of depressed patients, such as “I am such a loser,” or “I can’t go on.” After a series of such comments, the participants are asked to stop ruminating on the phrases and the thoughts they trigger. Researchers will measure how quickly subjects can disengage from negative thoughts, typically a difficult task for the depressed.
The process will be repeated for a control group that undergoes muscle relaxation training and depression education instead of MBCT. While it’s possible that patients in the control part of the study also will have reduced depressive symptoms, Desbordes said it should occur via different mechanisms in the brain, a difference that may be revealed by the scans. The work, which received funding from the National Center for Complementary and Integrative Health, has been underway since 2014 and is expected to last into 2019.
Desbordes said she wants to test one prevalent hypothesis about how MBCT works in depressed patients: that the training boosts body awareness in the moment, called interoception, which, by focusing their attention on the here and now, arms participants to break the cycle of self-rumination.
“We know those brain systems involved with interoception, and we know those involved with rumination and depression. I want to test, after taking MBCT, whether we see changes in these networks, particularly in tasks specifically engaging them,” Desbordes said.
Desbordes is part of a community of researchers at Harvard and its affiliated institutions that in recent decades has been teasing out whether and how meditation works.
In the 1970s, when transcendental meditation surged in popularity, Herbert Benson, a professor at Harvard Medical School and what was then Beth Israel Hospital, explored what he called “The Relaxation Response,” identifying it as the common, functional attribute of transcendental meditation, yoga, and other forms of meditation, including deep religious prayer. Benson described this response — which recent investigators say is not as common as he originally thought — as the opposite of the body’s adrenalin-charged “fight or flight” response, which was also identified at Harvard, by physiologist Walter Cannon Bradford in 1915.
Other MGH researchers also are studying the effects of meditation on the body, including Sara Lazar, who in 2012 used fMRI to show that the brains of subjects thickened after an eight-week meditation course. Work is ongoing at MGH’s Benson-Henry Institute; at HMS and Brigham and Women’s Hospital’s Osher Center for Integrative Medicine; at the Harvard-affiliated Cambridge Health Alliance, where Zev Schuman-Olivier directs the Center for Mindfulness and Compassion; and among a group of nearly a dozen investigators at Harvard and other Northeastern institutions, including Desbordes and Lazar, who are collaborating through the Mindfulness Research Collaborative.
Among the challenges researchers face is defining mindfulness itself. The word has come to describe a meditation-based practice whose aim is to increase one’s sense of being in the present, but it has also been used to describe a nonmeditative state in which subjects set aside their mental distractions to pay greater attention to the here and now, as in the work of Harvard psychologist Ellen Langer.
Another challenge involves sorting through the many variations of meditative practice.
Recent scientific exploration has largely focused on the secular practice of mindful meditation, but meditation is also a component of several ancient religious traditions, with variations. Even within the community practicing secular mindful meditation, there are variations that may be scientifically meaningful, such as how often one meditates and how long the sessions are. Desbordes herself has an interest in a variation called compassion meditation, whose aim is to increase caring for those around us.
Amid this variation, an eight-week mindfulness-based stress reduction course developed in the 1970s by Jon Kabat-Zinn at the University of Massachusetts Medical Center has become something of a clinical and scientific standard. The course involves weekly two- or 2½-hour group training sessions, 45 minutes of daily work on one’s own, and a daylong retreat. The mindfulness-based cognitive therapy used in Desbordes’ current work is a variation on that program and incorporates elements of cognitive behavioral therapy, which involves talk therapy effective in treating depression.
Ultimately, Desbordes said she’s interested in teasing out just what in mindful meditation can work against depression. If researchers can identify what elements are effective, the therapy may be refined to be more successful. Shapero is also interested in using the study to refine treatment. Since some patients benefit from mindfulness meditation and some do not, he’d like to better understand how to differentiate between the two.
“Once we know which ingredients are successful, we can do more of that and less, maybe, of the parts that are less effective,” Desbordes said.
Research funding includes the National Center for Complementary and Integrative Health.