tcd004 writes "MIT reports that the world is running out of fuel for our nuclear reactors due to production limitations and an aging infrastructure. Nuclear power has gained popularity as a carbon-free energy source in recent years, but Dr. Thomas Neff, a research affiliate at MIT's Center for International Studies, warned that fuel scarcity could drive up prices and kill the industry before it gets back on its feet. Passport has pulled together some interesting numbers: there are 440 reactors currently in operation and 82 new plants under construction. The demand for fuel has driven the price of uranium up more than 40% in the last few months — 900% over the last decade. You can follow the spot price for a pound of uranium here."
Iron Condor (964856) writes "The German magazine Der Spiegel has an article about the progress on the largest simulation of a part of a brain on the planet — with first interesting snippets and an outlook on the future (including, so the researchers hope, a simulation of a full human brain). Notable quote:
A project this ambitious would have been ridiculed a few years ago. "Today we have the computers we need," says biologist Henry Markram, 44, the project's director. "And we know enough to begin.""
memar76 writes "New Scientist has an article about an interesting new approach to developing conversational characters for virtual worlds by observing thousands of human players. The system attempts to harness the power of the masses, and learns by recording the words and actions of pairs of humans playing the roles of customers and waitresses in a virtual restaurant. This data will be used to teach AI characters how to react in any situation. The project website says that over 2,100 games have been played to date. This approach bears resemblance to commercial products from AiLive and Artificial Contender, but focuses on learning contextual dialogue rather than achieving competitive goals in sports or RPGs. Are these behavior capture systems the future of AI for games? The developer of The Restaurant Game project makes a convincing argument on his blog: "It is unrealistic to believe human AI programmers will ever account for the nuances of every possible scenario. Think about an analogy to animation — when animators want convincing human movement, they go straight to the source and use motion capture. Animators blend, layer, sequence captured animations to generate new motions beyond those recorded. Why should we believe that authoring AI behaviors is any different — perhaps we can get the most believable results by starting with the source material, and procedurally blending, layering, and sequencing behaviors as we see fit.""