Saltar para: Posts [1], Pesquisa e Arquivos [2]



 

How much can we afford to forget, if we train machines to remember?

Gene Tracy

(...)

Most of us no longer know how to grow the food we eat or build the homes we live in, after all. We don’t understand animal husbandry, or how to spin wool, or perhaps even how to change the spark plugs in a car. Most of us don’t need to know these things because we are members of what social psychologists call ‘transactive memory networks’.

We are constantly engaged in ‘memory transactions’ with a community of ‘memory partners’, through activities such as conversation, reading and writing. As members of these networks, most people no longer need to remember most things. 

(...)

What’s new, however, is that many of our memory partners are now smart machines. But an AI – such as Google search – is a memory partner like no other. It’s more like a memory ‘super-partner’, immediately responsive, always available. And it gives us access to a large fraction of the entire store of human knowledge.

 

Researchers have identified several pitfalls in the current situation. For one, our ancestors evolved within groups of other humans, a kind of peer-to-peer memory network. Yet information from other people is invariably coloured by various forms of bias and motivated reasoning. They dissemble and rationalise. They can be mistaken. We have learned to be alive to these flaws in others, and in ourselves. But the presentation of AI algorithms inclines many people to believe that these algorithms are necessarily correct and ‘objective’. Put simply, this is magical thinking.

 

The most advanced smart technologies today are trained through a repeated testing and scoring process, where human beings still ultimately sense-check and decide on the correct answers. Because machines must be trained on finite data-sets, with humans refereeing from the sidelines, algorithms have a tendency to amplify our pre-existing biases – about race, gender and more. An internal recruitment tool used by Amazon until 2017 presents a classic case: trained on the decisions of its internal HR department, the company found that the algorithm was systematically sidelining female candidates. If we’re not vigilant, our AI super-partners can become super-bigots.

 

A second quandary relates to the ease of accessing information. In the realm of the nondigital, the effort required to seek out knowledge from other people, or go to the library, makes it clear to us what knowledge lies in other brains or books, and what lies in our own head. But researchers have foundthat the sheer agility of the internet’s response can lead to the mistaken belief, encoded in later memories, that the knowledge we sought was part of what we knew all along.

 

A new kind of civilisation seems to be emerging, one rich in machine intelligence, with ubiquitous access points for us to join in nimble artificial memory networks. Even with implants, most of the knowledge we’d access would not reside in our ‘upgraded’ cyborg brains, but remotely – in banks of servers. In an eye-blink, from launch to response, each Google search nowtravels on average about 1,500 miles to a data centre and back, and uses about 1,000 computers along the way. But dependency on a network also means taking on new vulnerabilities. The collapse of any of the webs of relations that our wellbeing depends upon, such as food or energy, would be a calamity. Without food we starve, without energy we huddle in the cold. And it is through widespread loss of memory that civilisations are at risk of falling into a looming dark age.

(...) in an educational setting, unlike collaborative chess or medical diagnostics, the student is not yet a content expert. The AI as know-it-all memory partner can easily become a crutch, while producing students who think they can walk on their own.

 

publicado às 04:58


A automação do universo humano

por beatriz j a, em 17.12.14

 

 

 

Machines are replacing workers faster than economic expansion creates new manufacturing positions. As industrial robots become cheaper and more adept, the gap between lost and added jobs will almost certainly widen.” 

From an employer’s perspective, this makes sense. Machines are the perfect employees. They never get sick or complain or sexually harass their colleagues (at least not yet), and the occasional software upgrade is a lot cheaper than the health insurance and pension plan demanded by a human worker. And yet despite periodic fretting by economists, we’re oddly passive about the implications of this trend, no doubt because of our nation’s longstanding enthusiasm for technology. Carr quotes cognitive scientist Donald Norman, who has observed, “[T]he machine-centered viewpoint compares people to machines and finds us wanting, incapable of precise, repetitive, accurate actions.”

 

Rather than humanize the machines, we seem intent on making our human institutions more machinelike. 

 

 

publicado às 05:10


Os humanos, cada vez mais dispensáveis...

por beatriz j a, em 27.10.14

 

 

 

 

 

publicado às 22:04


Os 'pets' do fututro

por beatriz j a, em 03.09.13

 

 

 

 

theremina:sethturin:The LittleDog RobotThis is the more advanced version of this robot, created by the University of Southern California. The robot is completely autonomous and trained by machine learning algorithms. SQUEEEEEEEEEEEE

 

The LittleDog Robot

This is the more advanced version of this robot, created by the University of Southern California. The robot is completely autonomous and trained by machine learning algorithms



publicado às 15:09


no cabeçalho, pintura de Paul Béliveau. mail b.alcobia@sapo.pt

Subscrever por e-mail

A subscrição é anónima e gera, no máximo, um e-mail por dia.



Arquivo

  1. 2019
  2. J
  3. F
  4. M
  5. A
  6. M
  7. J
  8. J
  9. A
  10. S
  11. O
  12. N
  13. D
  14. 2018
  15. J
  16. F
  17. M
  18. A
  19. M
  20. J
  21. J
  22. A
  23. S
  24. O
  25. N
  26. D
  27. 2017
  28. J
  29. F
  30. M
  31. A
  32. M
  33. J
  34. J
  35. A
  36. S
  37. O
  38. N
  39. D
  40. 2016
  41. J
  42. F
  43. M
  44. A
  45. M
  46. J
  47. J
  48. A
  49. S
  50. O
  51. N
  52. D
  53. 2015
  54. J
  55. F
  56. M
  57. A
  58. M
  59. J
  60. J
  61. A
  62. S
  63. O
  64. N
  65. D
  66. 2014
  67. J
  68. F
  69. M
  70. A
  71. M
  72. J
  73. J
  74. A
  75. S
  76. O
  77. N
  78. D
  79. 2013
  80. J
  81. F
  82. M
  83. A
  84. M
  85. J
  86. J
  87. A
  88. S
  89. O
  90. N
  91. D
  92. 2012
  93. J
  94. F
  95. M
  96. A
  97. M
  98. J
  99. J
  100. A
  101. S
  102. O
  103. N
  104. D
  105. 2011
  106. J
  107. F
  108. M
  109. A
  110. M
  111. J
  112. J
  113. A
  114. S
  115. O
  116. N
  117. D
  118. 2010
  119. J
  120. F
  121. M
  122. A
  123. M
  124. J
  125. J
  126. A
  127. S
  128. O
  129. N
  130. D
  131. 2009
  132. J
  133. F
  134. M
  135. A
  136. M
  137. J
  138. J
  139. A
  140. S
  141. O
  142. N
  143. D
  144. 2008
  145. J
  146. F
  147. M
  148. A
  149. M
  150. J
  151. J
  152. A
  153. S
  154. O
  155. N
  156. D



Pesquisar

  Pesquisar no Blog

Edicoespqp.blogs.sapo.pt statistics