Chomsky's "Where AI went wrong" is often ignored by the mainstream AI community or dismissed. Peter Norvig's retort was poorly constructed, and showed that he didn't comprehend Chomsky's argument.
Machine Learning and AI are stuck in a rut and apparently these so called ML/AI experts know better. The Deep Learning (along with ML) craze is a hindrance to a true scientific theory of intelligence.
I don't see why we won't be able to reference common 3rd party libraries along with their version number and cache them across websites. Eventually a giant repo of common libraries will be built.
To explain that, you first need to understand what options are.
Options are a financial instrument where you can make bets on the market. To give you a simple example of an option, suppose you have a house worth 250k that you want to sell. You want to sell it within a year, but you are afraid that your house might lose value within that year (this is called the maturity). You might decide to instead sell an option contract for $5k, pricing the house at $250k which can be exercised within a year. What does this mean? This means that as a buyer, I pay $5k for this contract for the right to buy the underlying asset at $250k at any point within that year. The buyer is betting that the house value might appreciate beyond $250k, and the seller wants to protect themselves from the scenario that their house value depreciates. In general. options are all about betting what will happen in the future.
The problem now is in pricing your option contract. The earliest and most famous model of pricing comes from the Black Scholes model for pricing European options. With European options, you only exercise your option at the end of the maturity of the contract. However, with American options, things are a lot more complex, since the buy may decide to exercise their option at anytime within the maturity period. Moreover 99% of the industry trades with American options. For pricing American Options, the best method is to use a binomial or trinomial tree with many nodes (I will let you look up what those are).
The problem with the binomial and trinomial tree is that, it's very slow. You could lower the number of nodes in your tree, but then you lose accuracy. There are analytical solutions like Ju-Zhong which are very fast. Unfortunately such approximations tend to lose accuracy, and their error blows up. Finding an accurate approximation for pricing American options is a significant breakthrough for the industry. This problem has been eluding the best researchers in the field for decades.
What we have is an analytical solution to option pricing that is very fast, and highly accurate.
Why are we interested in option pricing? For one thing, it is a $500+ trillion dollar industry, and a lot of money is being lost in pricing errors. More importantly, if you are given an option contract, you can infer the implied volatility of the underlying asset (volatility is the degree of variation of the price of the underlying asset; could be stocks, real-estate etc.) . Since our solution is highly accurate and fast, we can extract the implied volatility in real time, along with the associated partial derivatives and make real time decision and assessments on the market.
Machine Learning and AI are stuck in a rut and apparently these so called ML/AI experts know better. The Deep Learning (along with ML) craze is a hindrance to a true scientific theory of intelligence.