Computers Ltd: What They Really Can't Do by David Harel

By David Harel

The pc has been hailed because the maximum innovation of the 20 th century, and there's no denying that those technological marvels have dramatically replaced our daily lives. they could fly airplanes and spaceships, direction hundreds of thousands of mobile calls at the same time, and play chess with the world's maximum avid gamers. yet how unlimited is the longer term for the pc? Will pcs someday be really clever, make scientific diagnoses, run businesses, compose tune, and fall in love? In pcs Ltd., David Harel, the best-selling writer of Algorithmics, illuminates probably the most primary but under-reported elements of computers--their inherent barriers. having a look in basic terms on the undesirable information that's confirmed, discussing obstacles that no quantities of undefined, software program, expertise, or assets can conquer, the ebook provides a worrying and provocative view of computing at the beginning of the twenty first century. Harel takes us on a desirable journey that touches on every thing from tiling difficulties and monkey puzzles to Monte Carlo algorithms and quantum computing, displaying simply how faraway from ideal desktops are, whereas shattering a number of the many claims made for those machines. He concludes that notwithstanding we may perhaps attempt for higher and higher issues in computing, we have to be real looking: desktops aren't omnipotent--far from it. Their limits are actual and the following to stick. in line with challenging evidence, mathematically confirmed and undeniable, desktops Ltd. deals a vividly written and infrequently fun examine the form of the longer term.

Show description

Read Online or Download Computers Ltd: What They Really Can't Do PDF

Best intelligence & semantics books

Cambrian Intelligence: The Early History of the New AI

This ebook is a suite of the "best" / so much mentioned Brooks papers. primarily it covers what's thought of the center of papers that bought behaviour dependent robotics rolling. just about all papers have seemed as magazine papers past and this is often simply a handy selection of those. For an individual engaged on cellular robotics those papers are a needs to.

Recent Advances in AI Planning: 4th European Conference on Planning, ECP'97, Toulouse, France, September 24 - 26, 1997, Proceedings

This booklet constitutes the refereed lawsuits of the 4th ecu convention on making plans, ECP'97, held in Toulouse, France, in September 1997. The 35 revised complete papers provided have been conscientiously reviewed and chosen from ninety submissions. the variety of themes coated spans all facets of present man made intelligence making plans, from theoretical and foundational issues to real making plans of platforms and purposes in numerous parts.

Artificial Intelligence: Its Scope and Limits

This sequence will contain monographs and collections of reports dedicated to the research and exploration of data, details, and knowledge­ processing platforms of every kind, regardless of even if human, (other) animal, or computer. Its scope is meant to span the whole variety of pursuits from classical difficulties within the philosophy of brain and philosophical psycholo­ gy via concerns in cognitive psychology and sociobiology (concerning the psychological functions of different species) to principles relating to man made in­ telligence and to machine technological know-how.

Application of Evolutionary Algorithms for Multi-objective Optimization in VLSI and Embedded Systems

This ebook describes how evolutionary algorithms (EA), together with genetic algorithms (GA) and particle swarm optimization (PSO) can be used for fixing multi-objective optimization difficulties within the quarter of embedded and VLSI method layout. Many complicated engineering optimization difficulties will be modelled as multi-objective formulations.

Additional resources for Computers Ltd: What They Really Can't Do

Sample text

G. ). In such a version, the rotations-forbidden constraint is redundant, and the bad news is exactly the same. s o m e t i m e s we c a n ' t d o i t 31 In way of illustration, consider Fig. 2, which shows three tile types and a 5 x 5 tiling. The reader will have no difficulty verifying that the pattern in the figure can be extended in all directions, to yield a tiling of any sized room whatsoever. As can also be seen, this tiled portion uses only the three available types, and it adheres to the color-matching rule.

34 comp u t e r s Ltd . Well, no, dear reader, you cannot. When we label a problem non­ computable or undecidable, we really and truly mean it. You can't solve it, and neither can anyone else, no matter how rich or patient or smart. Still, the claim does sound very strange if we don't restrict the basic operations. Surely, if anything is allowed, the following two­ step procedure solves the tiling problem: 1 . If the tile types in the input set T can tile rooms of any size, output 'Yes' and stop; 2.

The facts, however, are quite the other way around. I I I I If the available portion of the infinite plane is finite, the problem is trivially decidable, since only finitely many possible snakes can be positioned in a given finite area, and an algorithm can be easily designed continued on next page 48 c omp u t e r s Ltd . program verifi cation In Chapter 1 we discussed the need for algorithms and programs to be correct. Establishing the fact that a candidate program indeed solves the algorithmic problem you are working on is no easy feat.

Download PDF sample

Rated 4.50 of 5 – based on 38 votes