**Review From User :**

Edward Ashford Lee's Plato and the Nerd became a technical narrative respecting computational ability. Lee's narrative started with a differentiation between science and engineering and then delved into science for computer engineering. The narrative format captured Lee's technical acumen as well as conversational persona for this investigative look into computational power and limitations. One of Lee's favorite books inspired the title, Plato and the Nerd, and this title fit Lee's discourse on digital creationism. A majority of Lee's insight stemmed from a successful academic career. Plato and the Nerd included the fundamentals required for an appreciation of basic computational constraints. Bounds to computation had always been hardware as Lee illustrated during Plato and the Nerd. Hardware required immense resources for just 4 bit logic before today's 64 bit devices flourished. Cyberphysical systems blossomed because micro-fabrication of ALU's enabled the implementation of embedded software and portable hardware. Complex cyberphysical systems introduced uncertainty into computations, so Lee deduced the relevance of probabilistic models with simplified examples. Limits on probabilistic models appeared as their finiteness approached infinite, and Lee described the finitude for computation per Shannon's Law and the Bekenstein bound. Engineering's future thus departed independent and deterministic disciplines for hybridization and stochastics.

Engineering, from Lee's perspective, becomes the application of sciences for the solution of a valuable problem. Typical solutions of value pertain to business or social problems. Science, on the other hand, addresses any observable phenomena. An example of the difference appears where science includes Newton's Laws of Motion while engineering produces actuators such as motors, solenoids, pneumatics, and hydraulics. Science provides the laws and theories. Engineering represents the application process of scientific laws and theories to product development and lifecycle tasks. Lee's primary focus discusses the role of computational models in engineering. Physical prototypes arise from computational models known as analytical prototypes, and the quality of a computational model depends on programmatic sophistication as well as the computer hardware. Cukier and Schonberger adduce another aspect of computational modelling, the input data, yet Lee's discussion does not declare any number of inputs and deliberates instead on programs. A preference for functional languages has emerged as engineers provide programmable descriptions of systems not stepwise instructions, and the computers infer their consequent computations. The tools for computation have changed the acceptable applied sciences and engineering methods, and the outcome produces the engineering of models for a vast array of problem and solution variations.

Cyberphysical systems have become the union of several engineering fields: computer, mechanical, electrical, and software. Development of Lee's observations arise from the required computation involved in manifestation of physical objectives given specific mechanical, electrical, and computer hardware. Computer engineering empowers the nervous system of automated products for example servers, PLC's, communication networks, and microprocessors. A system's responsiveness to operator or environmental input is often the responsibility of computer components. Mechanical components of cyberphysical systems generate a product's interaction with the tangible or physical world based on computed instructions. Physicists have named these interactions work and heat transfer per the First Law of Thermodynamics, and mechanical engineers design mechanisms based on the processes of work and heat transfer. Turbines, refrigerators, and cars represent mechanisms where increased automation has required the involvement of on-board electronics and controls. Electrical engineers create nano scale SoC's for software engineers who embed intelligence into non Turing machines, so integration platforms such as IoT orchestrate multiple tasks. Proper integration of these mechanisms means hybridization of the traditional engineering schools into modernized fields: mechatronics, automation, and robotics exemplify new fields of engineering. Each new field requires subjects from the traditional mechanical, electrical, computer, and software schools for success.

Uncertainty appears inevitable for future computational models. Lee discusses deterministic examples of simple physical problems and then graduates the complexity of these examples. A simple example of an elastic collision between two pool balls becomes complicated once Lee introduces a third pool ball. Now, pool is a game with 15 balls plus 1 cue ball, and the first move in pool requires a collision amongst all 16 balls. If Lee observes inconsistencies with deterministic models for a collision amongst just three balls, then the deterministic computation of a sixteen ball collision seems fruitless and cumbersome. Bayesian probabilistic models resolve the most variability from deterministic models, and Lee prefers Bayesian models to frequentist approaches because Bayesian models do not require time. Time and any other continuous infinite set present too large of a data set for computation. Discussion of the Bekenstein bound grounds Lee's requirement for discrete finite sets because these sets permit a finite number of events for computation. Continuous datasets of genuine interest require discretization techniques such as a Fourier Transform. Discrete events with discrete and finite likelihood for a finite universe comprise set of computable events per Lee's rationale for a computational model of any desired cyberphysical system.

Edward Ashford Lee's Plato and the Nerd shall remain known for Lee's accessible explanations respecting computer engineering theory. Lee's differentiation of science and engineering will become important as the demand for scientific theory vs engineered solutions fluctuates with the available technologies. Scientific breakthroughs should precede engineering innovations because Lee defines engineering as applied science. Different fields of engineers might unite as proliferated cyberphysical developments require knowledge of complimentary disciplines. The limits of cyberphysics would enable accurate computation for solutions of finite Bayesian probabilistic models. Cyberphysics should increase the prevalence of automated or programmable non-Turing machines. Automated material handling systems as well as IoT devices may provide good examples of such non-Turing machines because their intended function will include more than just computation. Finite Bayesian probabilistic models can describe cyberphysical events to a maximum Bekenstein bound, and any beyond the Bekenstein bound would exceed the current scope of Lee's book. A relationship between entropy and digital information shall appear critical for investigation as cyberphysical system complexity increases, so increased quantum level experiments could yield results. Quantum computing may represent the norm for engineers as the boundaries of science and engineering expand because cubits will offer increased resolution and speed during computation.

**Media Size :** 11.7 MB