If you are looking for an answer to the question What is Artificial Intelligence? and you only have a minute, then here's the definition the Association for the Advancement of Artificial Intelligence offers on its home page: "the scientific understanding of the mechanisms underlying thought and intelligent behavior and their embodiment in machines."
However, if you are fortunate enough to have more than a minute, then please get ready to embark upon an exciting journey exploring AI (but beware, it could last a lifetime) …
Video Friday is your weekly selection of awesome robotics videos, collected by your Automaton bloggers. We'll also be posting a weekly calendar of upcoming robotics events for the next few months; here's what we have so far (send us your events!): Let us know if you have suggestions for next week, and enjoy today's videos. In case you missed it, NASA JPL landed a robot on Mars. What Thanksgiving is all about: robot arms and flamethrowers.
We derive a new intrinsic social motivation for multi-agent reinforcement learning (MARL), in which agents are rewarded for having causal influence over another agent's actions. Causal influence is assessed using counterfactual reasoning. The reward does not depend on observing another agent's reward function, and is thus a more realistic approach to MARL than taken in previous work. We show that the causal influence reward is related to maximizing the mutual information between agents' actions. We test the approach in challenging social dilemma environments, where it consistently leads to enhanced cooperation between agents and higher collective reward. Moreover, we find that rewarding influence can lead agents to develop emergent communication protocols. We therefore employ influence to train agents to use an explicit communication channel, and find that it leads to more effective communication and higher collective reward. Finally, we show that influence can be computed by equipping each agent with an internal model that predicts the actions of other agents. This allows the social influence reward to be computed without the use of a centralised controller, and as such represents a significantly more general and scalable inductive bias for MARL with independent agents.
Imagine one hundred years ago if farmers had access to huge volumes of information about the soil profile of their land, the varieties of crops they were growing, and even the fluctuations of their local climate. This kind of information could have prevented an environmental crisis like the Dust Bowl of the 1920s in the American Midwest. But even ten years ago, the idea that farmers could have access to this kind of information was unrealistic. For the team behind the CGIAR Platform for Big Data in Agriculture, farming is the next frontier for using artificial intelligence (AI) to efficiently solve complex problems. The team--which includes biologists, agronomists, nutritionists, and policy analysts working with data scientists--is using Big Data tools to create AI systems that can predict the potential outcomes of future scenarios for farmers.
For the many potential application areas where artificial intelligence can deliver breakthrough transformation, unlock efficiencies and augment human life, one of the most impactful areas – at least in the Indian context – will be the value it can bring to the field of agriculture. According to Indian Brand Equity Foundation, nearly 58% of India's population relies on agriculture as their primary source of livelihood. The total export of agricultural commodities for India is expected to hit $38.1 billion in FY18, making the country one of the top 15 exporters globally. However, agriculture in India is riddled with systemic problems, which AI is well placed to address. First, the traditionally unorganised agriculture sector continues to be difficult to organise due to our vast geography combined with our cultural and linguistic diversity.
The increased demand is adding to a big problem in the agricultural industry: a labor shortage that is unlikely to improve soon, according to farmers and researchers. It means berries that would otherwise be sold might instead rot in the field. Technologists think automation eventually could help producers pick specialty crops like berries, apples, peaches and snacking tomatoes, just as high-tech combines from companies such as Deere & Co. already are helping farmers of commodity crops harvest grains. Autonomous robots that run along tracks also are ferrying bins inside greenhouses, cutting down the walking workers have to do, growers and industry researchers say. But unlike in manufacturing, where artificial intelligence and computer vision power factory arms that move car parts or handle food in predetermined ways, agricultural fields pose a challenge for machines.
Hughes, Edward, Leibo, Joel Z., Phillips, Matthew G., Tuyls, Karl, Duéñez-Guzmán, Edgar A., Castañeda, Antonio García, Dunning, Iain, Zhu, Tina, McKee, Kevin R., Koster, Raphael, Roff, Heather, Graepel, Thore
Groups of humans are often able to find ways to cooperate with one another in complex, temporally extended social dilemmas. Models based on behavioral economics are only able to explain this phenomenon for unrealistic stateless matrix games. Recently, multi-agent reinforcement learning has been applied to generalize social dilemma problems to temporally and spatially extended Markov games. However, this has not yet generated an agent that learns to cooperate in social dilemmas as humans do. A key insight is that many, but not all, human individuals have inequity averse social preferences. This promotes a particular resolution of the matrix game social dilemma wherein inequity-averse individuals are personally pro-social and punish defectors. Here we extend this idea to Markov games and show that it promotes cooperation in several types of sequential social dilemma, via a profitable interaction with policy learnability. In particular, we find that inequity aversion improves temporal credit assignment for the important class of intertemporal social dilemmas. These results help explain how large-scale cooperation may emerge and persist.
When it comes to video games, violence sells. This year's E3, the biggest news event in the video game calendar, featured a lot of violent content, like every year before. It's saddening how much of gaming is defined solely in those terms – especially when more relaxing alternatives have being going through a mini-renaissance recently. Chill out games that focus on growing and relaxing rather than shooting or racing are not a rarity. They are not given the same exposure as their blockbuster counterparts, but they are plentiful and popular.
According to the UN, the world's population is estimated to increase by 29% to 9.8 billion in 2050. There is therefore an immense pressure on the world's agriculture sector to develop sustainable measures of increasing their output so as to meet the demands of the ever increasing population around the world. Agriculture has always been one of the core economic activities in Thailand, with over 40% of Thai workers employed in the industry. However, this sector contributes to only 10% of the economy and is on a decline. Fortunately, Thailand, having been backed by its government, is developing new technology to help transform its agriculture industry.
"One day the AIs are going to look back on us the same way we look at fossil skeletons on the plains of Africa. An upright ape living in dust with crude language and tools, all set for extinction." And so begins our fear of AI. Not to mention that HBO's Westworld and Channel 4's Humans fill our minds with futures that include self-aware beautiful android beings. There's also Sophia the Robot, who possesses an Instagram bio complete with, "Robots should have equal rights as humans."
If the predictions of the UN's Food and Agricultural Organization (FAO) are to be believed, the population of our blue planet is set to reach 9.2 billion by the year 2050. Currently, we are already touching the limits of available acreage for planting crops and breeding cattle. Estimates indicate that we only have as little as 4% available land for agriculture to potentially expand into. So, obviously, we will not be able to feed more mouths by adding more farmland. Instead, we need to process the land we have more efficiently in order to achieve higher yields.