Chat GPT: A Game-Changer

There has been much chatter that AI will likely take over our work and skillsets to make all our traditional learning and knowledge obsolete. Well, it looks like that is happening and will grow rapidly in this new year and will upend a great deal of the human resource need in the workplace. Knowledge workers, as well as the educational institutions that have afforded the formal training for them may be largely obsolete in five years :

There are two types of knowledge and of functioning: The type that everyone is trained in depends on memory – You are taught by your parents how to behave and in your education to study a bunch of information, and you later rely on your memory of these things and your association and inferences among them to guide you in how to innovate and operate in society, work, and in the world. But there is another type of knowledge that arises spontaneously, independent of our memories – it not only affords its own type of intelligence (like in being able to feel and ‘read’ people more deeply, for just one example), but also a spontaneity that makes life into more of a type of play than work. It is only available to us self-aware, conscious beings – machines and AI will never be able to replicate it. So as ChatGPT and similar types of AI systems evermore outperform us in the memory-based knowledge domain, this other more spontaneous, untrained type of knowingness/wisdom will most likely become all the more important to have. Think about it. Although Chat GPT has garnered the most attention, there are some who say that alternatives like GPT-3 are already more powerful.

I just used Chat GPT to ask a question that I wondered about in my earlier days in the physical sciences. This was my question:

“Symplectic geometry, and its connections to the Hamiltonian and Lagrangian formalisms, has been a powerful tool in classical mechanics, statistical mechanics, and in particle physics, but only when these systems are non-dissipative. What analogs or extensions exist for characterizing dissipative systems, such as the Navier-Stokes equations ?”

This is the answer it gave me in a few seconds:

“There are a few different ways to approach this problem, but one possibility is to use the theory of dissipative systems. This theory is based on the idea of a “generalized Poisson bracket,” which is a generalization of the Poisson bracket that is commonly used in symplectic geometry. The generalized Poisson bracket takes into account the dissipative terms in the system’s equations of motion, and can be used to derive generalized Hamiltonian and Lagrangian formalisms for dissipative systems.

One way to derive a generalized Poisson bracket is to start with the standard Poisson bracket and modify it to take into account dissipation. For example, if we have a system with a dissipative force F that is proportional to the velocity v, we can modify the Poisson bracket as follows:

[F,G] = ∂F/∂q * ∂G/∂p – ∂F/∂p * ∂G/∂q – γ * (p * ∂G/∂p + q * ∂G/∂q)

where q and p are the position and momentum variables, and γ is a coefficient that represents the strength of the dissipation.

Another approach is to use the theory of non-equilibrium thermodynamics, which is a branch of physics that deals with systems that are far from equilibrium. In this theory, the dissipation is described using a quantity called the “dissipation function,” which is used to derive equations of motion for the system.

There are also a number of other approaches that have been developed to study dissipative systems, including methods based on the theory of dynamical systems, the theory of nonlinear oscillations, and the theory of stochastic processes. Each of these approaches has its own strengths and limitations, and the choice of which one to use depends on the specific problem at hand.”