45 review for The Fifth Generation: Artificial Intelligence & Japan's Computer Challenge to the World

1. 

   5 out of 5

   Edward – May 22, 2014

   There is a lot of matter of fact repetition in the beginning chapters but it is interesting look back from a perspective of the state of computers three decades later and the great expectations at the time of publication. Good explanations of the basic issues associated with the development of AI. Some of the AI applications mentioned have come to fruition but not by means of AI. The authors were extremely optimistic and saw their book as a call to arms. The second half of the book is spent on be There is a lot of matter of fact repetition in the beginning chapters but it is interesting look back from a perspective of the state of computers three decades later and the great expectations at the time of publication. Good explanations of the basic issues associated with the development of AI. Some of the AI applications mentioned have come to fruition but not by means of AI. The authors were extremely optimistic and saw their book as a call to arms. The second half of the book is spent on belaboring the state of the U.S. economy, which hasn't improved in the interim, and how we are going to turn into a third rate nation militarily and economically if we don't start investing whole hog in the AI R&D program. The style of writing is somewhat odd because the authors refer to themselves in a detached manner. The authors do a good job in trying to dispel the myth that the Japanese are copycats lacking in original thinking; however, no attention was paid to it at the time based on my own experience. The last two chapters were excellent summaries on the possible effects that knowledge technology might have on humans. Of course the benefits of technology were expected to raise the entire world out of misery and poverty which hasn't happened. The world was transformed not because of Japan’s 5th Generation program, which failed, or because the U.S. didn’t follow suite but for more fundamental reasons; viz. globalization, the breakup of the Soviet Union, and corporate greed.

2. 

   5 out of 5

   Paco Nathan – Jul 03, 2016

   Hilarious.

3. 

   4 out of 5

   Rod Van Meter – Aug 10, 2020

   I finally read this book, 37 years after it was published at the early stages -- near peak hype -- of Japan's fifth generation computing project. I think understanding what happened with 5th gen (not today's 5G mobile networks!) is an important object lesson for both the hubris of early 1980s AI, and the dangers of large-scale, all-or-nothing, bet-the-house research projects. Modern American conservatives would say it's a lesson in why governments shouldn't "pick winners and losers". If you don't I finally read this book, 37 years after it was published at the early stages -- near peak hype -- of Japan's fifth generation computing project. I think understanding what happened with 5th gen (not today's 5G mobile networks!) is an important object lesson for both the hubris of early 1980s AI, and the dangers of large-scale, all-or-nothing, bet-the-house research projects. Modern American conservatives would say it's a lesson in why governments shouldn't "pick winners and losers". If you don't know the 5th gen KIPS (knowledge information processing systems) project, or FGCS, funded by MITI at the height of its own powers, you should. Japanese semiconductor and computer industries had a checkered record, from the point of view of the U.S., specializing in copying American designs for mainframes and peripherals, and even having lost important trade suits over theft of intellectual property. (Although it's not mentioned in this book, Intel at the time was intensely annoyed at Japanese manufacturers for taking over the DRAM market, though in the long run they may have done Intel a favor, forcing it to focus on its CPUs.) To step beyond that international image as a copycat, MITI created the FGCS project. Of course, that was unfair to a certain extent, as well; Japan had tremendous talent in the form of people like Kazuhiro Fuchi, Keio's own Aiso-sensei, and WIDE Project member Wada-sensei, all of the same generation as Gordon Bell and Dave Farber -- and Ed Feigenbaum, one of the book's authors and one of the founders of the field of expert systems. Fuchi was the one who got the call to lead FGCS. FGCS was supposed to move beyond mere databases (recall that RDMSes were still newish at the time), and create true knowledge processing: building but going beyond expert systems, rule-based AI -- what is sometimes called GOFAI, good old-fashioned AI. The project was *incredibly* ambitious, technologically. It required new semiconductor chips (and perhaps new fab processes), a new processor design, new programming languages, a new OS/runtime, and of course a lot of advances in applications to achieve the natural language recognition, image recognition and overall knowledge base structure. They were even aiming for automated parallelism in language processing, with several orders of magnitude gain in processing speed compared to mainframe and LISP machine-based performance of the day, measured in inferences per second. And all of this was to be achieved in a decade, making simple mainframes obsolete. MITI gathered a team of young hotshots around Fuchi, coming from the best Japanese companies and labs, although some perhaps reluctantly. There was substantial, and it turns out warranted, skepticism about whether Japan had all the right expertise to pull this off; in particular, there there was not a lot of expertise on, well, expert systems. And that's pretty much where the story stood at the time of the writing of the book... Given the title, one might be forgiven for assuming this is a book about Japan and about the FGCS project itself. Instead, only the middle third of the book is really about the project, and only about a half-dozen Japanese people are mentioned by name. The first third of the book is about the history of AI and expert systems in particular, a reasonable and necessary introduction for the lay reader who may not be familiar with the field. (I found it enlightening history, from today's perspective.) Then the last third is about international reaction to the founding of the project, fears that Japan would outpace the U.S. and U.K. (and Germany and France, which make cameos), and become the "Saudi Arabia of knowledge". Naturally, this book sounds a call to arms for the U.S., and discusses Feigenbaum's vision for an equivalent project. From a 2020 perspective, there are a number of significant eye-rollers, what could easily be called thinly veiled racism or at least jingoism for a sustained American dominance in the field. A big chunk of the last third focuses on what knowledge systems can do. The vision hear sounds, well, much like today's Google -- easily searchable data, good NLP, even image recognition. It's actually a fascinating read, comparing that vision with what we have today. We may not have gotten a Prolog-programmed KIPS system, but what we have is pretty good.

4. 

   5 out of 5

   Jamie –

5. 

   5 out of 5

   Les Johnson –

6. 

   5 out of 5

   John Watson –

7. 

   4 out of 5

   Kaye –

8. 

   5 out of 5

   BookSwim.com Book Rental Online –

9. 

   5 out of 5

   Steve Hendricks –

10. 

    5 out of 5

    Fernando –

11. 

    5 out of 5

    Damian Tiamo –

12. 

    5 out of 5

    Subhajit Das –

13. 

    4 out of 5

    Phil Wolff –

14. 

    5 out of 5

    K –

15. 

    4 out of 5

    Peter Castine –

16. 

    4 out of 5

    Mark –

17. 

    4 out of 5

    Alan –

18. 

    5 out of 5

    David Schwan –

19. 

    4 out of 5

    Prospero –

20. 

    5 out of 5

    Max Luebbe –

21. 

    4 out of 5

    Lars Fischer –

22. 

    5 out of 5

    Will –

23. 

    4 out of 5

    Des Small –

24. 

    5 out of 5

    Ben Damer –

25. 

    4 out of 5

    Mark Lacy –

26. 

    4 out of 5

    Steve –

27. 

    5 out of 5

    B.B. –

28. 

    5 out of 5

    Eric Leuliette –

29. 

    4 out of 5

    Sergey –

30. 

    4 out of 5

    Nick –

31. 

    4 out of 5

    Philomath –

32. 

    5 out of 5

    .LAG –

33. 

    4 out of 5

    Joseph –

34. 

    4 out of 5

    Qaim Hassan –

35. 

    5 out of 5

    Rana –

36. 

    4 out of 5

    Aqib Butt –

37. 

    5 out of 5

    Carlos –

38. 

    4 out of 5

    Almosawy Ahmed –

39. 

    5 out of 5

    Afshan Mohtashim –

40. 

    4 out of 5

    Joyce Rostan –

41. 

    4 out of 5

    Arpit Mohanty –

42. 

    5 out of 5

    Lisha –

43. 

    4 out of 5

    Rahul Rajpurohit –

44. 

    4 out of 5

    Muzafarali –

45. 

    4 out of 5

    Madushan –