The paper attempts to unify scientifically the two extremes in the form of Duality of Randomness & Causality, Predictability & Uncertainty in Market like Wave-Particle duality scientifically in Nature. And how that should be leveraged in Financial Investment & Risk Management Strategies in Real World Practical Applications.Conventional Financial/Factor Investing has been studied and practiced from an economic perspective logic for a long time through the application of Conventional flawed Popular Statistical Tools. Factor Investing (Often Misunderstood by the Modelers) affecting trillions of dollars in investment globally seems to lack the proper Scientific base. This paper aims to check the conceptual Scientific validity of Financial Investment, Factor Investing, Valuation, Conventional Risk Management Tools, Black Swan , Tail Risk, Convexity, Conventional Statistical Tools e.g. Linear Regression, and Popular Risk measurement metrics e.g. Sharpe Ratio, Max Drawdown etc. , Historically important Puzzle e.g. Equity Risk Premium. Then it further explains the proposes the scientific foundation of Convexity, Causal Laws of Motion in Markets, Causal Based Randomized Factor Investing for Portfolio Management, and how Financial Investment & Risk Management should be studied and practiced in a Scientific way in the Real World rather than based on many economic theories without proper Scientific validation. It also talks about how Value should be redefined Scientifically in Real World Investment unlike Conventionally flawed definition . It looks scientifically at the foundational issues with the Conventional Statistical /Mathematical /AI/ML tools e.g. Non-Stationarity ,Linear Regression etc. applied in Real World Financial Investment and tries to make it scientific in improved way. It also proposes the new Causal Based Randomized Learning Approach for ML in Investment in Real World.This is basically a Conceptual Paper that also looks at the foundation of Conventional Mathematical & Statistical Tools applied in Investment & Risk domains e.g. Monte Carlo Simulations Probability theory, Expectations etc and how that need to be evolved in Real World Investment driven by Quantum Human Behavior Dynamics. In a nutshell, It explores how financial investment and risk management/taking should be done scientifically in forward direction of time rather backward direction of time as usually happens in most statistical based models in financial investment and risk areas.