Showing posts with label 3D occupancy grid. Show all posts
Showing posts with label 3D occupancy grid. Show all posts

Artificial Intelligence - Who Is Hans Moravec?

 




Hans Moravec(1948–) is well-known in the computer science community as the long-time head of Carnegie Mellon University's Robotics Institute and an unashamed techno logical optimist.

For the last twenty-five years, he has studied and produced artificially intelligent robots at the CMU lab, where he is still an adjunct faculty member.

Moravec spent almost 10 years as a research assistant at Stanford University's groundbreaking Artificial Intelligence Lab before coming to Carnegie Mellon.

Moravec is also noted for his paradox, which states that, contrary to popular belief, it is simple to program high-level thinking skills into robots—as with chess or Jeopardy!—but difficult to transmit sensorimo tor agility.

Human sensory and motor abilities have developed over millions of years and seem to be easy, despite their complexity.

Higher-order cognitive abilities, on the other hand, are the result of more recent cultural development.

Geometry, stock market research, and petroleum engineering are examples of disciplines that are difficult for people to learn but easier for robots to learn.

"The basic lesson of thirty-five years of AI research is that the hard issues are simple, and the easy ones are hard," writes Steven Pinker of Moravec's scientific career.

Moravec built his first toy robot out of scrap metal when he was eleven years old, and his light-following electronic turtle and a robot operated by punched paper tape earned him two high school science fair honors.

He proposed a Ship of Theseus-like analogy for the viability of artificial brains while still in high school.

Consider replacing a person's human neurons one by one with precisely manufactured equivalents, he said.

When do you think human awareness will vanish? Is anybody going to notice? Is it possible to establish that the person is no longer human? Later in his career, Moravec would suggest that human knowledge and training might be broken down in the same manner, into subtasks that machine intelligences could take over.

Moravec's master's thesis focused on the development of a computer language for artificial intelligence, while his PhD research focused on the development of a robot that could navigate obstacle courses utilizing spatial representation methods.

The area of interest (ROI) in a scene was identified by these robot vision systems.

Moravec's early computer vision robots were extremely sluggish by today's standards, taking around five hours to go from one half of the facility to the other.

To measure distance and develop an internal picture of physical impediments in the room, a remote computer carefully analysed continuous video-camera images recorded by the robot from various angles.

Moravec finally developed 3D occupancy grid technology, which allowed a robot to create an awareness of a cluttered area in a matter of seconds.

Moravec's lab took on a new challenge by converting a Pontiac TransSport minivan into one of the world's first road-ready autonomous cars.

The self-driving minivan reached speeds of up to 60 miles per hour.

DANTE II, a robot capable of going inside the crater of an active volcano on Mount Spurr in Alaska, was also constructed by the CMU Robotics Institute.

While DANTE II's immediate aim was to sample harmful fumarole gases, a job too perilous for humans, it was also planned to demonstrate technologies for robotic expeditions to distant worlds.

The volcanic explorer robot used artificial intelligence to navigate the perilous, boulder-strewn terrain on its own.

Because such rovers produced so much visual and other sensory data that had to be analyzed and managed, Moravec believes that experience with mobile robots spurred the development of powerful artificial intelligence and computer vision methods.

For the National Aeronautics and Space Administration (NASA), Moravec's team built fractal branching ultra-dexterous robots ("Bush robots") in the 1990s.

These robots, which were proposed but never produced due to the lack of necessary manufacturing technologies, comprised of a branching hierarchy of dynamic articulated limbs, starting with a main trunk and splitting down into smaller branches.

As a result, the Bush robot would have "hands" at all scales, from macroscopic to tiny.

The tiniest fingers would be nanoscale in size, allowing them to grip very tiny objects.

Moravec said the robot would need autonomy and depend on artificial intelligence agents scattered throughout the robot's limbs and branches because to the intricacy of manipulating millions of fingers in real time.

He believed that the robots may be made entirely of carbon nanotube material, employing the quick prototyping technology known as 3D printers.

Moravec believes that artificial intelligence will have a significant influence on human civilization.

To stress the role of AI in change, he coined the concept of the "landscape of human capability," which physicist Max Tegmark has later converted into a graphic depiction.

Moravec's picture depicts a three-dimensional environment in which greater altitudes reflect more challenging jobs in terms of human difficulty.

The point where the swelling waters meet the shore reflects the line where robots and humans both struggle with the same duties.

Art, science, and literature are now beyond of grasp for an AI, but the sea has already defeated mathematics, chess, and the game Go.

Language translation, autonomous driving, and financial investment are all on the horizon.

More controversially, in two popular books, Mind Children (1988) and Robot: Mere Machine to Transcendent Mind (1989), Moravec engaged in future conjecture based on what he understood of developments in artificial intelligence research (1999).

In 2040, he said, human intellect will be surpassed by machine intelligence, and the human species would go extinct.

Moravec evaluated the functional equivalence between 50,000 million instructions per second (50,000 MIPS) of computer power and a gram of brain tissue and came up with this figure.

He calculated that home computers in the early 2000s equaled only an insect's nervous system, but that if processing power doubled every eighteen months, 350 million years of human intellect development could be reduced to just 35 years of artificial intelligence advancement.

He estimated that a hundred million MIPS would be required to create human-like universal robots.

Moravec refers to these sophisticated robots as our "mind children" in the year 2040.

Humans, he claims, will devise techniques to delay biological civilization's final demise.

Moravec, for example, was the first to anticipate what is now known as universal basic income, which is delivered by benign artificial superintelligences.

In a completely automated society, a basic income system would provide monthly cash payments to all individuals without any type of employment requirement.

Moravec is more concerned about the idea of a renegade automated corporation breaking its programming and refusing to pay taxes into the human cradle-to-grave social security system than he is about technological unemployment.

Nonetheless, he predicts that these "wild" intelligences will eventually control the universe.

Moravec has said that his books Mind Children and Robot may have had a direct impact on the last third of Stanley Kubrick's original screenplay for A.I. Artificial Intelligence (later filmed by Steven Spielberg).

Moravecs, on the other hand, are self-replicating devices in the science fiction books Ilium and Olympos.

Moravec defended the same physical fundamentalism he expressed in his high school thoughts throughout his life.

He contends in his most transhumanist publications that the only way for humans to stay up with machine intelligences is to merge with them by replacing sluggish human cerebral tissue with artificial neural networks controlled by super-fast algorithms.

In his publications, Moravec has blended the ideas of artificial intelligence with virtual reality simulation.


He's come up with four scenarios for the development of consciousness.

(1) human brains in the physical world, 

(2) a programmed AI implanted in a physical robot, 

(3) a human brain immersed in a virtual reality simulation, and 

(4) an AI functioning inside the boundaries of virtual reality All of them are equally credible depictions of reality, and they are as "real" as we believe them to be.


Moravec is the creator and chief scientist of the Pittsburgh-based Seegrid Corporation, which makes autonomous Robotic Industrial Trucks that can navigate warehouses and factories without the usage of automated guided vehicle systems.

A human trainer physically pushes Seegrid's vehicles through a new facility once.

The robot conducts the rest of the job, determining the most efficient and safe pathways for future journeys, while the trainer stops at the appropriate spots for the truck to be loaded and unloaded.

Seegrid VGVs have transported over two million production miles and eight billion pounds of merchandise for DHL, Whirlpool, and Amazon.

Moravec was born in the Austrian town of Kautzen.

During World War II, his father was a Czech engineer who sold electrical products.

When the Russians invaded Czechoslovakia in 1944, the family moved to Austria.

In 1953, his family relocated to Canada, where he now resides.

Moravec earned a bachelor's degree in mathematics from Acadia University in Nova Scotia, a master's degree in computer science from the University of Western Ontario, and a doctorate from Stanford University, where he worked with John McCarthy and Tom Binford on his thesis.

The Office of Naval Study, the Defense Advanced Research Projects Agency, and NASA have all supported his research.

Elon Musk (1971–) is an American businessman and inventor.

Elon Musk is an engineer, entrepreneur, and inventor who was born in South Africa.

He is a dual citizen of South Africa, Canada, and the United States, and resides in California.

Musk is widely regarded as one of the most prominent inventors and engineers of the twenty-first century, as well as an important influencer and contributor to the development of artificial intelligence.

Despite his controversial personality, Musk is widely regarded as one of the most prominent inventors and engineers of the twenty-first century and an important influencer and contributor to the development of artificial intelligence.

Musk's business instincts and remarkable technological talent were evident from an early age.

By the age of 10, he had self-taught himself how program computers, and by the age of twelve, he had produced a video game and sold the source code to a computer maga zine.

Musk has included allusions to some of his favorite novels in SpaceX's Falcon Heavy rocket launch and Tesla's software since he was a youngster.

Musk's official schooling was centered on economics and physics rather than engineering, interests that are mirrored in his subsequent work, such as his efforts in renewable energy and space exploration.

He began his education at Queen's University in Canada, but later transferred to the University of Pennsylvania, where he earned bachelor's degrees in Economics and Physics.

Musk barely stayed at Stanford University for two days to seek a PhD in energy physics before departing to start his first firm, Zip2, with his brother Kimbal Musk.

Musk has started or cofounded many firms, including three different billion-dollar enterprises: SpaceX, Tesla, and PayPal, all driven by his diverse interests and goals.

• Zip2 was a web software business that was eventually purchased by Compaq.

• X.com: an online bank that merged with PayPal to become the online payments corporation PayPal.

• Tesla, Inc.: an electric car and solar panel maker • SpaceX: a commercial aircraft manufacturer and space transportation services provider (via its subsidiarity SolarCity) • Neuralink: a neurotechnology startup focusing on brain-computer connections • The Boring Business: an infrastructure and tunnel construction corporation • OpenAI: a nonprofit AI research company focused on the promotion and development of friendly AI Musk is a supporter of environmentally friendly energy and consumption.

Concerns over the planet's future habitability prompted him to investigate the potential of establishing a self-sustaining human colony on Mars.

Other projects include the Hyperloop, a high-speed transportation system, and the Musk electric jet, a jet-powered supersonic electric aircraft.

Musk sat on President Donald Trump's Strategy and Policy Forum and Manufacturing Jobs Initiative for a short time before stepping out when the US withdrew from the Paris Climate Agreement.

Musk launched the Musk Foundation in 2002, which funds and supports research and activism in the domains of renewable energy, human space exploration, pediatric research, and science and engineering education.

Musk's effect on AI is significant, despite his best-known work with Tesla and SpaceX, as well as his contentious social media pronouncements.

In 2015, Musk cofounded the charity OpenAI with the objective of creating and supporting "friendly AI," or AI that is created, deployed, and utilized in a manner that benefits mankind as a whole.

OpenAI's objective is to make AI open and accessible to the general public, reducing the risks of AI being controlled by a few privileged people.

OpenAI is especially concerned about the possibility of Artificial General Intelligence (AGI), which is broadly defined as AI capable of human-level (or greater) performance on any intellectual task, and ensuring that any such AGI is developed responsibly, transparently, and distributed evenly and openly.

OpenAI has had its own successes in taking AI to new levels while staying true to its goals of keeping AI friendly and open.

In June of 2018, a team of OpenAI-built robots defeated a human team in the video game Dota 2, a feat that could only be accomplished through robot teamwork and collaboration.

Bill Gates, a cofounder of Microsoft, praised the achievement on Twitter, calling it "a huge milestone in advancing artificial intelligence" (@BillGates, June 26, 2018).

Musk resigned away from the OpenAI board in February 2018 to prevent any conflicts of interest while Tesla advanced its AI work for autonomous driving.

Musk became the CEO of Tesla in 2008 after cofounding the company in 2003 as an investor.

Musk was the chairman of Tesla's board of directors until 2018, when he stepped down as part of a deal with the US Securities and Exchange Commission over Musk's false claims about taking the company private.

Tesla produces electric automobiles with self-driving capabilities.

Tesla Grohmann Automation and Solar City, two of its subsidiaries, offer relevant automotive technology and manufacturing services and solar energy services, respectively.

Tesla, according to Musk, will reach Level 5 autonomous driving capabilities in 2019, as defined by the National Highway Traffic Safety Administration's (NHTSA) five levels of autonomous driving.

Tes la's aggressive development with autonomous driving has influenced conventional car makers' attitudes toward electric cars and autonomous driving, and prompted a congressional assessment of how and when the technology should be regulated.

Musk is widely credited as a key influencer in moving the automotive industry toward autonomous driving, highlighting the benefits of autonomous vehicles (including reduced fatalities in vehicle crashes, increased worker productivity, increased transportation efficiency, and job creation) and demonstrating that the technology is achievable in the near term.

Tesla's autonomous driving code has been created and enhanced under the guidance of Musk and Tesla's Director of AI, Andrej Karpathy (Autopilot).

The computer vision analysis used by Tesla, which includes an array of cameras on each car and real-time image processing, enables the system to make real-time observations and predictions.

The cameras, as well as other exterior and internal sensors, capture a large quantity of data, which is evaluated and utilized to improve Autopilot programming.

Tesla is the only autonomous car maker that is opposed to the LIDAR laser sensor (an acronym for light detection and ranging).

Tesla uses cameras, radar, and ultrasonic sensors instead.

Though academics and manufacturers disagree on whether LIDAR is required for fully autonomous driving, the high cost of LIDAR has limited Tesla's rivals' ability to produce and sell vehicles at a pricing range that allows a large number of cars on the road to gather data.

Tesla is creating its own AI hardware in addition to its AI programming.

Musk stated in late 2017 that Tesla is building its own silicon for artificial-intelligence calculations, allowing the company to construct its own AI processors rather than depending on third-party sources like Nvidia.

Tesla's AI progress in autonomous driving has been marred by setbacks.

Tesla has consistently missed self-imposed deadlines, and serious accidents have been blamed on flaws in the vehicle's Autopilot mode, including a non-injury accident in 2018, in which the vehicle failed to detect a parked firetruck on a California freeway, and a fatal accident in 2018, in which the vehicle failed to detect a pedestrian outside a crosswalk.

Neuralink was established by Musk in 2016.

With the stated objective of helping humans to keep up with AI breakthroughs, Neuralink is focused on creating devices that can be implanted into the human brain to better facilitate communication between the brain and software.

Musk has characterized the gadgets as a more efficient interface with computer equipment, while people now operate things with their fingertips and voice commands, directives would instead come straight from the brain.

Though Musk has made major advances to AI, his pronouncements regarding the risks linked with AI have been apocalyptic.

Musk has called AI "humanity's greatest existential danger" and "the greatest peril we face as a civilisation" (McFarland 2014).

(Morris 2017).

He cautions against the perils of power concentration, a lack of independent control, and a competitive rush to acceptance without appropriate analysis of the repercussions.

While Musk has used colorful terminology such as "summoning the devil" (McFarland 2014) and depictions of cyborg overlords, he has also warned of more immediate and realistic concerns such as job losses and AI-driven misinformation campaigns.

Though Musk's statements might come out as alarmist, many important and well-respected figures, including as Microsoft cofounder Bill Gates, Swedish-American scientist Max Tegmark, and the late theoretical physicist Stephen Hawking, share his concern.

Furthermore, Musk does not call for the cessation of AI research.

Instead, Musk supports for responsible AI development and regulation, including the formation of a Congressional committee to spend years studying AI with the goal of better understanding the technology and its hazards before establishing suitable legal limits.


~ Jai Krishna Ponnappan

Find Jai on Twitter | LinkedIn | Instagram


You may also want to read more about Artificial Intelligence here.



See also: 


Superintelligence; Technological Singularity; Workplace Automation.



References & Further Reading:


Moravec, Hans. 1988. Mind Children: The Future of Robot and Human Intelligence. Cambridge, MA: Harvard University Press.

Moravec, Hans. 1999. Robot: Mere Machine to Transcendent Mind. Oxford, UK: Oxford University Press.

Moravec, Hans. 2003. “Robots, After All.” Communications of the ACM 46, no. 10 (October): 90–97.

Pinker, Steven. 2007. The Language Instinct: How the Mind Creates Language. New York: Harper.




What Is Artificial General Intelligence?

Artificial General Intelligence (AGI) is defined as the software representation of generalized human cognitive capacities that enables the ...