Artificial Intelligence News for August 22 2017

Artificial intelligence

Artificial intelligence – or AI – is the science of programming machines or computers to reproduce human processes, like learning and decision making.

Artificial Intelligence

A new statistical analysis Q&A website launched. While the proposal for a statistical analysis Q&A website on area51 is taking it’s time, and the website is still collecting people who will commit to it,Joseph Turian, who seems a nice guy from his various comments online, seem to feel this website is not what the community needs and that we shouldn’t hold up on our questions for the website to go online. Joseph is pushing with all his might his newest creation “MetaOptimize QA”, a StackOverFlow like website for: machine learning, natural language processing, artificial intelligence, text analysis, information retrieval, search, data mining, statistical modeling, and data visualization. Is this new website better then the area51 website? Will all the people go to just one of the two websites. My own suggestion is to try both websites and let time tell. Didn’t we just start pushing forward another statistical Q&A website two weeks ago? I am talking about the Stack Exchange Q&A site proposal: Statistical Analysis. Should we wait for Stack Exchange offer for a new website to start? Or should we start using MetaOptimize? On the other hand, I do believe that the stackoverflow people have more experience in handling such websites then Joseph. I can very easily trust them to do regular database backups, share the websites database dumps with the general community, smoothly test and upgrade to provide new features, and generally speaking perform in a more experienced way with the online Q&A community. Tal.: Let’s start with the easy question, should I worry about technical issues in the website? Tal.: Me fear that this might fork the people in the field to two websites, instead of bringing them together. If the Stack Exchange website where to launch today, I would probably focus on using it and hint to the site for MetaOptimize. If the stack exchange version of the website where to start in a few weeks, I would probably sit on the fence and see if people are using it. What if the website where to start in a week, what then? I have no clue.
more…

The Facts, Fiction, and Future of Artificial Intelligence

While artificial intelligence isn’t exactly new, it has become a very hot topic as of late. The story of modern AI dates back to 1956, when revered computer scientist John McCarthy came up with the name “Artificial Intelligence”, which then became the study of using computer systems to perform tasks that require human intelligence. Ironically, the artificial intelligence you see in video games isn’t the same AI John McCarthy and others envisioned for this exciting concept. If you’re like most of us, you probably got your first taste of artificial intelligence from Hollywood. From one flick to the next, these fictional machines portray both the emotion, and lack of emotion associated with human beings. Movie critics are hailing Transcendence as one of the best Hollywood depictions of AI. The 2014 film tells the story of a Will Caster, a dedicated artificial intelligence researcher whose life’s mission is to develop a sentient system that integrates all knowledge known to man in a single machine. Not to be too much of a spoiler, but Caster eventually becomes the focal point of his own experiment as an artificial being with the consciousness of mankind and limitless wisdom is unleashed on the world. Though often compelling from an entertainment standpoint, Hollywood feeds the myths and misconceptions that surround artificial intelligence. A machine faces even greater challenges in this scenario due to its lack of cultural experiences, hardships, and everyday human interactions. Even the most skilled of smart machines still depend on man to write their automation capabilities, perform repairs, and handle general maintenance. Without us, the most advanced AI machines would be non-functional sooner or later. Cinema cheese aside, the potential of artificial intelligence fits within the realm of reality and is currently playing out as we speak. Companies you may be more familiar with, namely search giant Google, is leveraging it to make the world smarter, more productive, and efficient through the aid of machines with brain-like processors. Others fear the threat of machines putting more humans out of work, out-thinking us in the battle of wits, and simply falling into the wrong hands.
more…

Edu Introduction As computer games become more complex and consumers demand more sophisticated computer controlled agents, developers are required to place a greater emphasis on the artificial intelligence aspects of their games. Our experience developing intelligent air combat agents for DARPA training exercises, described in John Laird’s lecture at the 1998 Computer Game Developer’s Conference, suggested that many principles and techniques from the research community are applicable to games. Simple agents have been created for Quake II and Descent 3 and complex agents, that share a large amount of knowledge, are under development. From a game development perspective, the main goal of the Soar/Games project is to make games more enjoyable by making the agents in games more intelligent and realistic. An AI engine will also make the development of intelligent agents for games easier by providing a common inference machine and general knowledge base that can be easily applied to new games. Context specific agents ensure that their actions are consistent with past sensor information and the agent’s past actions. Perhaps the most common approach to building intelligent agents in games is to use C code to implement the AI with a large number of nested if and case statements. A second class of inference machines common in games use scripted sequences of actions to generate the agent’s behavior. At specific points in the game or when the agent senses certain conditions, the inference machine begins to execute one of the scripts stored in its knowledge base. On a 300 MHz Pentium II machine Soar can handle 6-10 agents Top Level Goals Attack Retreat Explore Collect Power-ups Tactics Circle-Strafe Charge Camp Snipe Pop out Chase Behaviors Find hidden location Select weapon Wait for target Shoot target Figure 1: A portion of a sample operator hierarchy for an action game such as Quake II or Descent 3. One of the common complaints about game AI is that the agents are allowed to cheat. Requiring the intelligent agents to use the same sensor information, follow the same rules and use the same actions as the human players eliminates cheating and results in realistic agents. Descent 3’s flying agents might circle-strafe in three dimensions, while Quake agents would circle-strafe in only two dimensions. A speech interface to cooperative agents in Quake II is being developed which will allow a human player to act as an officer, giving voice commands coordinating the actions of a platoon of intelligent agent soldiers.
more…

Kai-Fu Lee on artificial intelligence in China

A visionary venture capitalist and engineer on the present and future of Chinese tech. K.ai-Fu Lee is one of the most prominent figures in Chinese technology. He founded China’s noted early-stage venture capital firm Sinovation Ventures after launching and heading up Google’s China operations during their years of growth from 2005 to 2009. Born in Taiwan and educated at Columbia and Carnegie Mellon, Kai-Fu had an early career in Silicon Valley, including a stint as principal research scientist at Apple. Microsoft brought him to Beijing in 1998 to set up a research division, as he has seen the rise of the Chinese internet from its earliest days. Kai-Fu has more than 50 million fans on the social media platform Weibo and is a much-loved public speaker and author. He is perhaps most admired for his gutsy investing in Chinese startup companies: Sinovation puts money into startup companies in their riskiest early years or even months. Kai-Fu founded it in 2009, at least half a decade before the world began to take Chinese innovation seriously. Now Kai-Fu is turning his attention to artificial intelligence, and he spoke to Kaiser and Jeremy about it for this podcast at – of all places – the Trump International Tower in midtown New York City. Jiayang Fan from the New Yorker was finishing off an interview as they arrived, and she stayed for the chat. The discussion ranges from new technologies that are coming from Chinese engineers to the inexorable rise of AI and how it will change the way we live, work, and think. Jiayang: Abacus: Small Enough to Jail, a documentary on the only bank in America prosecuted for mortgage fraud, which brings the characters of the Abacus Federal Savings of Chinatown in New York to life. Kai-Fu: An “Anti-recommendation” against all sci-fi movies except one: Robot & Frank. Kaiser: “Friends Like These: How a famed Chinese dissident got caught up in America’s culture wars,” the 2013 Reuters profile of the political kerfuffle in the U.S. over blind human rights lawyer Chen Guangcheng.
more…