Slashdot Mirror
Scientists Crowdfund The Theory of Everything (cphpost.dk)
einar.petersen writes: Danish scientists are seeking to fund their research on the theory of everything in a rather unconventional way, namely via crowdfunding. The two researchers have launched a campaign that as of writing is 55% funded....
"Einstein spent the last 30 years of his life searching for an answer to the deepest question about the universe: does a fundamental principle, that governs all of reality, exist...?" reads their Indiegogo page. "In 2013 we, the theoretical physicist Jesper Moller Grimstrup and the mathematician Johannes Aastrup, discovered a simple mathematical principle, which we believe could be exactly what Einstein was searching for." One Danish newspaper jokes that the mathematician and theoretical physicist "are now offering mere mortals a chance to get in on the action."
"Einstein spent the last 30 years of his life searching for an answer to the deepest question about the universe: does a fundamental principle, that governs all of reality, exist...?" reads their Indiegogo page. "In 2013 we, the theoretical physicist Jesper Moller Grimstrup and the mathematician Johannes Aastrup, discovered a simple mathematical principle, which we believe could be exactly what Einstein was searching for." One Danish newspaper jokes that the mathematician and theoretical physicist "are now offering mere mortals a chance to get in on the action."
I've been wondering whether AI systems may advance science @ some point. I mean: not just as a tool with a human at the control knobs & interpreting results, but by itself as the 'entity' doing the advancing.
Some significant advances have been made not through heaps of grunt work, but when great minds like Einstein did their thing. Seeing patterns in their mind that no-one else saw. Sadly, such great minds are rare. And have a limited lifespan - of which a big part is spent learning the subject matter. And no matter how genius, with hard limits on the # of grey cells that can be thrown at the problem.
Artificial intelligent systems don't have such flesh-and-blood limitations: these can effectively be built at will, any size, optimized for specific problems sets, etc. Lately computerized systems have beat humans at increasingly complex tasks. Sometimes using brute force. Sometimes by looking at a problem from many angles at once. Fed with enough data, 'seeing' connections somehow that even experts in the field might overlook.
Regardless how it works exactly, fact is you might say that for some problems, we've built AI systems that are more capable than a "genius" human at finding solutions. Would it be hard to imagine that @ some point, an AI system might spit out a new formula, discover some as-of-yet-unseen regularity in scientific data, or find a path to unify as-of-yet-non-unified scientific theories?
Exciting times...