The ability to predict future prices sounds all well and good. However, In the face of the 'unpredictable' failures of global economics which even the experts could not predict, I don't know how a machine could be any other than a trial-and-error system. Still, a project to look out for.
Thank you for your question.
This is in fact we have been asking ourselves many times. But then we can across the academic research of Didier Sornette at ETH Zurich (Switzerland,
www.ethz.ch) and several others.
Professor Sornette has worked on predicting economic crises for around 10 years at ETHZ, and he is also a former professional physicist and chief risk office at Bank of America. He has also given a tech talk on using statistics to predict crises in the future:
https://www.ted.com/talks/didier_sornette_how_we_can_predict_the_next_financial_crisis[/u]]https://www.ted.com/talks/didier_sornette_how_we_can_predict_the_next_financial_crisisOf course, AI as powerful as NVIDIA's DGX-1 is now has only been available since May 2017 (for ETHZ, as a VIP customer and No. ~15 best university in the world, a bit earlier). These new machines are
able to be trained, and in fact
can train themselves to accomplish tasks. You might have seen recent videos of robotic dogs
https://www.youtube.com/watch?v=M8YjvHYbZ9w. Some of those
robotic dogs train themselves to walk and to act.
TESLA cars are accumulating driving data and will learn to drive themselves eventually (powered by NVIDIA processors). With AI predictions of future prices today of cryptocurrencies, houses and stocks, it is the same - the new AI supercomputers are very expensive but they can train themselves to solve economic problems ever better. Yes,
better than economists - just like a car can travel faster than a human. Economists are perhaps too close to politicians, while
computers consider only objective facts.
The biggest financial firms in the world have owned supercomputers since
Renaissance Technologies (
https://en.wikipedia.org/wiki/Renaissance_Technologies) in 1982. Renaissance and other firms make $ billions per year in profits - so much so that its founder funded Donald Trump's election victory. Until now, only the wealthiest, secretive customers were able to benefit from information advantage that supercomputers can deliver.
The planned
House Panda supercomputer will be more powerful than most supercomputers Renaissance and other hedge funds/banks have been running in recent memory. For a relatively low monthly subscription fee (starting at
99HPT tokens), everybody will be able to have access to information and future predictions that have so far been only available to the very wealthiest, and mostly in several western countries.
We are happy that House Panda has started at exactly the right time - when
AI, crypto and the potential customer base are all experiencing exponential innovation and growth.