Just as electricity transformed the way industries functioned in the past century, artificial intelligence — the science of programming cognitive abilities into machines — has the power to substantially change society in the next 100 years. AI is being harnessed to enable such things as home robots, robo-taxis and mental health chatbots to make you feel better.
A startup is developing robots with AI that brings them closer to human level intelligence. Already, AI has been embedding itself in daily life — such as powering the brains of digital assistants Siri and Alexa. It lets consumers shop and search online more accurately and efficiently, among other tasks that people take for granted.
“AI is the new electricity,” said Andrew Ng, co-founder of Coursera and an adjunct Stanford professor who founded the Google Brain Deep Learning Project, in a keynote speech at the AI Frontiers conference that was held this past weekend in Silicon Valley. “About 100 years ago, electricity transformed every major industry. AI has advanced to the point where it has the power to transform” every major sector in coming years. And even though there’s a perception that AI was a fairly new development, it has actually been around for decades, he said. But it is taking off now because of the ability to scale data and computation.
Ng said most of the value created through AI today has been through supervised learning, in which an input of X leads to Y. But there have been two major waves of progress: One wave leverages deep learning to enable such things as predicting whether a consumer will click on an online ad after the algorithm gets some information about him. The second wave came when the output no longer has to be a number or integer but things like speech recognition, a sentence structure in another language or audio. For example, in self-driving cars, the input of an image can lead to an output of the positions of other cars on the road.
Indeed, deep learning — where a computer learns from datasets to perform functions, instead of just executing specific tasks it was programmed to do — was instrumental in achieving human parity in speech recognition, said Xuedong Huang, who led the team at Microsoft on the historic achievement in 2016 when their system booked a 5.9% error rate, the same as a human transcriptionist. “Thanks to deep learning, we were able to reach human parity after 20 years,” he said at the conference. The team has since lowered the error rate even more, to 5.1%.
“We have cheap motors, sensors, batteries, plastics and processors … why don’t we have Rosie?”–Dileep George
The Rise of Digital Assistants
Starting in 2010, the quality of speech recognition began to improve for the industry, eventually leading to the creation of Siri and Alexa. “Now, you almost take it for granted,” Ng said. That’s not all; speech is expected to replace touch-typing for input, said Ruhi Sarikaya, director of Amazon Alexa. The key to greater accuracy is to understand the context. For example, if a person asks Alexa what he should do for dinner, the digital assistant has to assess his intent. Is he asking Alexa to make a restaurant reservation, order food or find a recipe? If he asks Alexa to find ‘Hunger Games,’ does he want the music, video or audiobook?
And what’s next for the digital assistant is an even more advanced undertaking — to understand “meaning beyond words,” said Dilek Hakkani-Tur, research scientist at Google. For example, if the user uses the words “later today,” it could mean 7 p.m. to 9 p.m. for dinner or 3 p.m. to 5 p.m. for meetings. This next level up also calls for more complex and lively conversations, multi-domain tasks and interactions beyond domain boundaries, she said. Moreover, Hakkani-Tur said, digital assistants should be able to do things such as easily read and summarize emails.
After speech, ‘computer vision’ — or the ability of computers to recognize images and categorize them — was the next to leap, speakers said. With many people uploading images and video, it became cumbersome to add metadata to all content as a way to categorize them. Facebook built an AI to understand and categorize videos at scale called Lumos, said Manohar Paluri, a research lead at the company. Facebook uses Lumos to do data collection of, for example, fireworks images and videos. The platform can also use people’s poses to identify a video, such as categorizing people lounging around on couches as hanging out.
“Her job is to bring a spot of life to your home. She provides entertainment — she can play music, podcasts, audiobooks.”–Kaijen Hsiao
What’s critical is to ascertain the primary semantic content of the uploaded video, added Rahul Sukthankar, head of video understanding at Google. And to help the computer correctly identify what’s in the video — for example, whether professionals or amateurs are dancing — his team mines YouTube for similar content that AI can learn from, such as having a certain frame rate for non-professional content. Sukthankar adds that a promising direction for future research is to do computer training using videos. So if a robot is shown a video of a person pouring cereal into a bowl at multiple angles, it should learn by watching.
At Alibaba, AI is used to boost sales. For example, shoppers of its Taobao e-commerce site can upload a picture of a product they would like to buy, like a trendy handbag sported by a stranger on the street, and the website will come up with handbags for sale that come closest to the photo. Alibaba also uses augmented reality/virtual reality to make people see and shop from stores like Costco. On its Youku video site, which is similar to YouTube, Alibaba is working on a way to insert virtual 3D objects into people’s uploaded videos, as a way to increase revenue. That’s because many video sites struggle with profitability. “YouTube still loses money,” said Xiaofeng Ren, a chief scientist at Alibaba.
Rosie and the Home Robot
But with all the advances in AI, it’s still no match for the human brain. Vicarious is a startup that aims to close the gap by developing human level intelligence in robots. Co-founder Dileep George said that the components are there for smarter robots. “We have cheap motors, sensors, batteries, plastics and processors … why don’t we have Rosie?” He was referring to the multipurpose robot maid in the 1960s space-age cartoon The Jetsons. George said the current level of AI is like what he calls the “old brain,” similar to the cognitive ability of rats. The “new brain” is more developed such as what’s seen in primates and whales.
George said the “old brain” AI gets confused when small inputs are changed. For example, a robot that can play a video game goes awry when the colors are made just 2% brighter. “AI today is not ready,” he said. Vicarious uses deep learning to get the robot closer to human cognitive ability. In the same test, a robot with Vicarious’s AI kept playing the game even though the brightness had changed. Another thing that confuses “old brain” AI is putting two objects together. People can see two things superimposed on each other, such as a coffee mug partly obscuring a vase in a photo, but robots mistake it for one unidentified object. Vicarious, which counts Facebook CEO Mark Zuckerberg as an investor, aims to solve such problems.
The intelligence inside Kuri, a robot companion and videographer meant for the home, is different. Kaijen Hsiao, chief technology officer of creator Mayfield Robotics, said there is a camera behind the robot’s left eye that gathers video in HD. Kuri has depth sensors to map the home and uses images to improve navigation. She also has pet and person detection features so she can smile or react when they are around. Kuri has place recognition as well, so she will remember she has been to a place before even if the lighting has changed, such as the kitchen during the day or night. Moment selection is another feature of the robot, which lets her recognize similar videos she records — such as dad playing with the baby in the living room — and eliminates redundant ones.
“Her job is to bring a spot of life to your home. She provides entertainment — she can play music, podcasts, audiobooks. You can check your home from anywhere,” Hsiao said. Kuri is the family’s videographer, going around the house recording so no one is left out. The robot will curate the videos and show the best ones. For this, Kuri uses vision and deep learning algorithms. “Her point is her personality … [as] an adorable companion,” Hsiao said. Kuri will hit the market in December at $799.
“About 100 years ago, electricity transformed every major industry. AI has advanced to the point where it has the power to transform” every major sector in coming years.–Andrew Ng
Business Response to AI
The U.S. and China lead the world in investments in AI, according to James Manyika, chairman and director of the McKinsey Global Institute. Last year, AI investment in North America ranged from $15 billion to $23 billion, Asia (mainly China) was $8 billion to $12 billion, and Europe lagged at $3 billion to $4 billion. Tech giants are the primary investors in AI, pouring in between $20 billion and $30 billion, with another $6 billion to $9 billion from others, such as venture capitalists and private equity firms.
Where did they put their money? Machine learning took 56% of the investments with computer vision second at 28%. Natural language garnered 7%, autonomous vehicles was at 6% and virtual assistants made up the rest. But despite the level of investment, actual business adoption of AI remains limited, even among firms that know its capabilities, Manyika said. Around 40% of firms are thinking about it, 40% experiment with it and only 20% actually adopt AI in a few areas.
The reason for such reticence is that 41% of companies surveyed are not convinced they can see a return on their investment, 30% said the business case isn’t quite there and the rest said they don’t have the skills to handle AI. However, McKinsey believes that AI can more than double the impact of other analytics and has the potential to materially raise corporate performance.
There are companies that get it. Among sectors leading in AI are telecom and tech companies, financial institutions and automakers. Manyika said these early adopters tend to be larger and digitally mature companies that incorporate AI into core activities, focus on growth and innovation over cost savings and enjoy the support of C-suite level executives. The slowest adopters are companies in health care, travel, professional services, education and construction. However, as AI becomes widespread, it’s a matter of time before firms get on board, experts said.
Join The Discussion
2 Comments So Far
Anumakonda Jagadeesh
Interesting.
AI takeover refers to a hypothetical scenario in which artificial intelligence (AI) becomes the dominant form of intelligence on Earth, with computers or robots effectively taking control of the planet away from the human race. Possible scenarios include replacement of the entire human workforce, takeover by a superintelligent AI, and the popular notion of a robot uprising. Some public figures, such as Stephen Hawking and Elon Musk, have advocated research into precautionary measures to ensure future superintelligent machines remain under human control.[1] Robot rebellions have been a major theme throughout science fiction for many decades though the scenarios dealt with by science fiction are generally very different from those of concern to scientists.
Advantages of superhuman intelligence over humans
An Artifming abilities as a competent artificial intelligence researcher, would be able to modify its own source code and increase its own intelligence. If its self-reprogramming leads to its getting even better at being able to reprogram itself, the result could be a recursive intelligence explosion where it would rapidly leave human intelligence far behind.
• Technology research: A machine with superhuman scientific research abilities would be able to beat the human research community to milestones such as nanotechnology or advanced biotechnology. If the advantage becomes sufficiently large (for example, due to a sudden intelligence explosion), an AI takeover becomes trivial. For example, a superintelligent AI might design self-replicating bots that initially escape detection by diffusing throughout the world at a low concentration. Then, at a prearranged time, the bots multiply into nanofactories that cover every square foot of the Earth, producing nerve gas or deadly target-seeking mini-drones.
• Strategizing: A superintelligence might be able to simply outwit human opposition.
• Social manipulation: A superintelligence might be able to recruit human support, or covertly incite a war between humans.
• Economic productivity: As long as a copy of the AGI could produce more economic wealth than the cost of its hardware, individual humans would have an incentive to voluntarily allow the AGI to run a copy of itself on their systems.
• Hacking: A superintelligence could find new exploits in computers connected to the Internet, and spread copies of itself onto those systems, or might steal money to finance its plans.
Sources of AGI advantage
A computer program that faithfully emulates a human brain, or that otherwise runs algorithms that are equally powerful as the human brain’s algorithms, could still become a “speed superintelligence” if it can think many orders of magnitude faster than a human, due to being made of silicon rather than flesh, or due to optimization focusing on increasing the speed of the AGI. Biological neurons operate at about 200 Hz, whereas a modern microprocessor operates at a speed of about 2,000,000,000 Hz. Human axons carry action potentials at around 120 m/s, whereas computer signals travel near the speed of light.
A network of human-level intelligences designed to network together and share complex thoughts and memories seamlessly, able to collectively work as a giant unified team without friction, or consisting of trillions of human-level intelligences, would become a “collective superintelligence”.
More broadly, any number of qualitative improvements to a human-level AGI could result in a “quality superintelligence”, perhaps resulting in an AGI as far above us in intelligence as humans are above non-human apes. The number of neurons in a human brain is limited by cranial volume and metabolic constraints; in contrast, you can add components to a supercomputer until it fills up its entire warehouse. An AGI need not be limited by human constraints on working memory, and might therefore be able to intuitively grasp more complex relationships than humans can. An AGI with specialized cognitive support for engineering or computer programming would have an advantage in these fields, compared with humans who evolved no specialized mental modules to specifically deal with those domains. Unlike humans, an AGI can spawn off copies of itself and tinker with its copies’ source code to attempt to further improve its algorithms.[
Possibility of unfriendly AI preceding friendly AI
Is strong AI inherently dangerous?
A significant problem is that unfriendly artificial intelligence is likely to be much easier to create than friendly AI. While both require large advances in recursive optimisation process design, friendly AI also requires the ability to make goal structures invariant under self-improvement (or the AI could transform itself into something unfriendly) and a goal structure that aligns with human values and does not automatically destroy the human race. An unfriendly AI, on the other hand, can optimize for an arbitrary goal structure, which does not need to be invariant under self-modification.[
The sheer complexity of human value systems makes it very difficult to make AI’s motivations human-friendly. Unless moral philosophy provides us with a flawless ethical theory, an AI’s utility function could allow for many potentially harmful scenarios that conform with a given ethical framework but not “common sense”. According to Eliezer Yudkowsky, there is little reason to suppose that an artificially designed mind would have such an adaptation.
Necessity of conflict
For an AI takeover to be inevitable, it has to be postulated that two intelligent species cannot pursue mutually the goals of coexisting peacefully in an overlapping environment—especially if one is of much more advanced intelligence and much more powerful. While an AI takeover is thus a possible result of the invention of artificial intelligence, a peaceful outcome is not necessarily impossible.
The fear of cybernetic revolt is often based on interpretations of humanity’s history, which is rife with incidents of enslavement and genocide. Such fears stem from a belief that competitiveness and aggression are necessary in any intelligent being’s goal system. However, such human competitiveness stems from the evolutionary background to our intelligence, where the survival and reproduction of genes in the face of human and non-human competitors was the central goal.[18] In fact, an arbitrary intelligence could have arbitrary goals: there is no particular reason that an artificially intelligent machine (not sharing humanity’s evolutionary context) would be hostile—or friendly—unless its creator programs it to be such and it is not inclined or capable of modifying its programming. But the question remains: what would happen if AI systems could interact and evolve (evolution in this context means self-modification or selection and reproduction) and need to compete over resources, would that create goals of self-preservation? AI’s goal of self-preservation could be in conflict with some goals of humans.
Some scientists dispute the likelihood of cybernetic revolts as depicted in science fiction such as The Matrix, claiming that it is more likely that any artificial intelligence powerful enough to threaten humanity would probably be programmed not to attack it. This would not, however, protect against the possibility of a revolt initiated by terrorists, or by accident. Artificial General Intelligence researcher Eliezer Yudkowsky has stated on this note that, probabilistically, humanity is less likely to be threatened by deliberately aggressive AIs than by AIs which were programmed such that their goals are unintentionally incompatible with human survival or well-being (as in the film I, Robot and in the short story “The Evitable Conflict”). Steve Omohundro suggests that present-day automation systems are not designed for safety and that AIs may blindly optimize narrow utility functions (say, playing chess at all costs), leading them to seek self-preservation and elimination of obstacles, including humans who might turn them off.
Another factor which may negate the likelihood of an AI takeover is the vast difference between humans and AIs in terms of the resources necessary for survival. Humans require a “wet,” organic, temperate, oxygen-laden environment while an AI might thrive essentially anywhere because their construction and energy needs would most likely be largely non-organic. With little or no competition for resources, conflict would perhaps be less likely no matter what sort of motivational architecture an artificial intelligence was given, especially provided with the superabundance of non-organic material resources in, for instance, the asteroid belt. This, however, does not negate the possibility of a disinterested or unsympathetic AI artificially decomposing all life on earth into mineral components for consumption or other purposes.
Other scientists point to the possibility of humans upgrading their capabilities with bionics and/or genetic engineering and, as cyborgs, becoming the dominant species in themselves.
Criticism and counterarguments
Advantages of humans over superhuman intelligence
If a superhuman intelligence is a deliberate creation of human beings, theoretically its creators could have the foresight to take precautions in advance. In the case of a sudden “intelligence explosion”, effective precautions will be extremely difficult; not only would its creators have little ability to test their precautions on an intermediate intelligence, but the creators might not even have made any precautions at all, if the advent of the intelligence explosion catches them completely by surprise.
Boxing
An AGI’s creators would have two important advantages in preventing a hostile AI takeover: first, they could choose to attempt to “keep the AI in a box”, and deliberately limit its abilities. The tradeoff in boxing is that the creators presumably built the AGI for some concrete purpose; the more restrictions they place on the AGI, the less useful the AGI will be to its creators. (At an extreme, “pulling the plug” on the AGI makes it useless, and is therefore not a viable long-term solution.) A sufficiently strong superintelligence might find unexpected ways to escape the box, for example by social manipulation, or by providing the schematic for a device that ostensibly aids its creators but in reality brings about the AGI’s freedom, once built.
Instilling positive values
The second important advantage is that an AGI’s creators can theoretically attempt to instill human values in the AGI, or otherwise align the AGI’s goals with their own, thus preventing the AGI from wanting to launch a hostile takeover. However, it is not currently known, even in theory, how to guarantee this.
Warnings
Physicist Stephen Hawking, Microsoft founder Bill Gates and SpaceX founder Elon Musk have expressed concerns about the possibility that AI could develop to the point that humans could not control it, with Hawking theorizing that this could “spell the end of the human race”.[20] Stephen Hawking said in 2014 that “Success in creating AI would be the biggest event in human history. Unfortunately, it might also be the last, unless we learn how to avoid the risks.” Hawking believes that in the coming decades, AI could offer “incalculable benefits and risks” such as “technology outsmarting financial markets, out-inventing human researchers, out-manipulating human leaders, and developing weapons we cannot even understand.” In January 2015, Nick Bostrom joined Stephen Hawking,Max Tegmark, Elon Musk, Lord Martin Rees, Jaan Tallinn, and numerous AI researchers, in signing the Future of Life Institute’s open letter speaking to the potential risks and benefits associated with artificial intelligence(Source: Wikipedia).
Dr.A.Jagadeesh Nellore(AP),India
Alan Wells
“such as a coffee mug partly obscuring a vase in a photo, but robots mistake it for one unidentified object” This is an essential difference between humans and animals. A dog sees a man, a horse, and a man-on-a-horse. Three objects.