vidtheque.

read-only demo

Videos Ib5GBkD555M

Harness Engineering is not Enough: Why Software Factories Fail — Dex Horthy, HumanLayer

index_state ready data_status ok

AI Engineer· published 2026-07-23· 0:19:17· en-US· indexed 2026-08-10 19:41

Open on YouTube

Scene timeline

  1. Shot 0, 0:00 to 0:03, 1 of 1 keyframes kept
  2. Shot 1, 0:03 to 0:05, 1 of 1 keyframes kept
  3. Shot 2, 0:05 to 0:12, 1 of 1 keyframes kept
  4. Shot 3, 0:12 to 0:13, 1 of 1 keyframes kept
  5. Shot 4, 0:13 to 0:14, 1 of 1 keyframes kept
  6. Shot 5, 0:14 to 0:16, 1 of 1 keyframes kept
  7. Shot 6, 0:16 to 0:18, 1 of 1 keyframes kept
  8. Shot 7, 0:18 to 0:19, 0 of 1 keyframes kept
  9. Shot 8, 0:19 to 0:20, 1 of 1 keyframes kept
  10. Shot 9, 0:20 to 0:29, 1 of 1 keyframes kept
  11. Shot 10, 0:29 to 0:32, 1 of 1 keyframes kept
  12. Shot 11, 0:32 to 0:34, 1 of 1 keyframes kept
  13. Shot 12, 0:34 to 0:39, 1 of 1 keyframes kept
  14. Shot 13, 0:39 to 0:52, 1 of 1 keyframes kept
  15. Shot 14, 0:52 to 0:55, 1 of 1 keyframes kept
  16. Shot 15, 0:55 to 1:02, 1 of 1 keyframes kept
  17. Shot 16, 1:02 to 1:05, 1 of 1 keyframes kept
  18. Shot 17, 1:05 to 1:13, 1 of 1 keyframes kept
  19. Shot 18, 1:13 to 1:16, 1 of 1 keyframes kept
  20. Shot 19, 1:16 to 1:21, 1 of 1 keyframes kept
  21. Shot 20, 1:21 to 1:26, 1 of 1 keyframes kept
  22. Shot 21, 1:26 to 1:30, 1 of 1 keyframes kept
  23. Shot 22, 1:30 to 1:33, 1 of 1 keyframes kept
  24. Shot 23, 1:33 to 1:36, 1 of 1 keyframes kept
  25. Shot 24, 1:36 to 1:41, 1 of 1 keyframes kept
  26. Shot 25, 1:41 to 1:47, 1 of 1 keyframes kept
  27. Shot 26, 1:47 to 1:56, 1 of 1 keyframes kept
  28. Shot 27, 1:56 to 2:02, 1 of 1 keyframes kept
  29. Shot 28, 2:02 to 2:07, 1 of 1 keyframes kept
  30. Shot 29, 2:07 to 2:14, 1 of 1 keyframes kept
  31. Shot 30, 2:14 to 2:15, 1 of 1 keyframes kept
  32. Shot 31, 2:15 to 2:20, 1 of 1 keyframes kept
  33. Shot 32, 2:20 to 2:31, 1 of 1 keyframes kept
  34. Shot 33, 2:31 to 2:39, 1 of 1 keyframes kept
  35. Shot 34, 2:39 to 2:48, 1 of 1 keyframes kept
  36. Shot 35, 2:48 to 2:52, 1 of 1 keyframes kept
  37. Shot 36, 2:52 to 2:56, 1 of 1 keyframes kept
  38. Shot 37, 2:56 to 3:00, 1 of 1 keyframes kept
  39. Shot 38, 3:00 to 3:17, 1 of 1 keyframes kept
  40. Shot 39, 3:17 to 3:35, 1 of 1 keyframes kept
  41. Shot 40, 3:35 to 3:38, 1 of 1 keyframes kept
  42. Shot 41, 3:38 to 3:41, 0 of 1 keyframes kept
  43. Shot 42, 3:41 to 3:51, 1 of 1 keyframes kept
  44. Shot 43, 3:51 to 3:53, 1 of 1 keyframes kept
  45. Shot 44, 3:53 to 4:19, 1 of 1 keyframes kept
  46. Shot 45, 4:19 to 4:44, 1 of 1 keyframes kept
  47. Shot 46, 4:44 to 5:10, 0 of 1 keyframes kept
  48. Shot 47, 5:10 to 5:12, 0 of 1 keyframes kept
  49. Shot 48, 5:12 to 5:14, 1 of 1 keyframes kept
  50. Shot 49, 5:14 to 5:45, 1 of 1 keyframes kept
  51. Shot 50, 5:45 to 5:47, 0 of 1 keyframes kept
  52. Shot 51, 5:47 to 5:50, 1 of 1 keyframes kept
  53. Shot 52, 5:50 to 5:52, 1 of 1 keyframes kept
  54. Shot 53, 5:52 to 5:53, 1 of 1 keyframes kept
  55. Shot 54, 5:53 to 5:54, 1 of 1 keyframes kept
  56. Shot 55, 5:54 to 5:55, 1 of 1 keyframes kept
  57. Shot 56, 5:55 to 5:56, 1 of 1 keyframes kept
  58. Shot 57, 5:56 to 5:58, 1 of 1 keyframes kept
  59. Shot 58, 5:58 to 6:00, 1 of 1 keyframes kept
  60. Shot 59, 6:00 to 6:03, 1 of 1 keyframes kept
  61. Shot 60, 6:03 to 6:06, 1 of 1 keyframes kept
  62. Shot 61, 6:06 to 6:31, 1 of 1 keyframes kept
  63. Shot 62, 6:31 to 6:33, 1 of 1 keyframes kept
  64. Shot 63, 6:33 to 6:59, 0 of 1 keyframes kept
  65. Shot 64, 6:59 to 7:01, 1 of 1 keyframes kept
  66. Shot 65, 7:01 to 7:06, 1 of 1 keyframes kept
  67. Shot 66, 7:06 to 7:24, 1 of 1 keyframes kept
  68. Shot 67, 7:24 to 7:34, 1 of 1 keyframes kept
  69. Shot 68, 7:34 to 7:36, 1 of 1 keyframes kept
  70. Shot 69, 7:36 to 7:50, 1 of 1 keyframes kept
  71. Shot 70, 7:50 to 7:56, 0 of 1 keyframes kept
  72. Shot 71, 7:56 to 8:17, 1 of 1 keyframes kept
  73. Shot 72, 8:17 to 8:23, 1 of 1 keyframes kept
  74. Shot 73, 8:23 to 8:27, 1 of 1 keyframes kept
  75. Shot 74, 8:27 to 8:36, 1 of 1 keyframes kept
  76. Shot 75, 8:36 to 8:53, 1 of 1 keyframes kept
  77. Shot 76, 8:53 to 8:55, 1 of 1 keyframes kept
  78. Shot 77, 8:55 to 9:00, 1 of 1 keyframes kept
  79. Shot 78, 9:00 to 9:07, 1 of 1 keyframes kept
  80. Shot 79, 9:07 to 9:12, 1 of 1 keyframes kept
  81. Shot 80, 9:12 to 9:24, 1 of 1 keyframes kept
  82. Shot 81, 9:24 to 9:27, 1 of 1 keyframes kept
  83. Shot 82, 9:27 to 9:29, 1 of 1 keyframes kept
  84. Shot 83, 9:29 to 9:32, 1 of 1 keyframes kept
  85. Shot 84, 9:32 to 9:36, 1 of 1 keyframes kept
  86. Shot 85, 9:36 to 9:44, 1 of 1 keyframes kept
  87. Shot 86, 9:44 to 9:49, 1 of 1 keyframes kept
  88. Shot 87, 9:49 to 10:10, 1 of 1 keyframes kept
  89. Shot 88, 10:10 to 10:12, 1 of 1 keyframes kept
  90. Shot 89, 10:12 to 10:20, 1 of 1 keyframes kept
  91. Shot 90, 10:20 to 10:29, 1 of 1 keyframes kept
  92. Shot 91, 10:29 to 10:48, 1 of 1 keyframes kept
  93. Shot 92, 10:48 to 11:05, 1 of 1 keyframes kept
  94. Shot 93, 11:05 to 11:13, 1 of 1 keyframes kept
  95. Shot 94, 11:13 to 11:19, 1 of 1 keyframes kept
  96. Shot 95, 11:19 to 11:29, 1 of 1 keyframes kept
  97. Shot 96, 11:29 to 11:34, 1 of 1 keyframes kept
  98. Shot 97, 11:34 to 11:46, 1 of 1 keyframes kept
  99. Shot 98, 11:46 to 12:02, 1 of 1 keyframes kept
  100. Shot 99, 12:02 to 12:07, 1 of 1 keyframes kept
  101. Shot 100, 12:07 to 12:20, 1 of 1 keyframes kept
  102. Shot 101, 12:20 to 12:24, 1 of 1 keyframes kept
  103. Shot 102, 12:24 to 12:27, 1 of 1 keyframes kept
  104. Shot 103, 12:27 to 12:31, 1 of 1 keyframes kept
  105. Shot 104, 12:31 to 12:38, 1 of 1 keyframes kept
  106. Shot 105, 12:38 to 12:41, 0 of 1 keyframes kept
  107. Shot 106, 12:41 to 12:44, 1 of 1 keyframes kept
  108. Shot 107, 12:44 to 12:48, 1 of 1 keyframes kept
  109. Shot 108, 12:48 to 12:57, 1 of 1 keyframes kept
  110. Shot 109, 12:57 to 13:02, 1 of 1 keyframes kept
  111. Shot 110, 13:02 to 13:11, 1 of 1 keyframes kept
  112. Shot 111, 13:11 to 13:15, 1 of 1 keyframes kept
  113. Shot 112, 13:15 to 13:17, 1 of 1 keyframes kept
  114. Shot 113, 13:17 to 13:24, 1 of 1 keyframes kept
  115. Shot 114, 13:24 to 13:29, 1 of 1 keyframes kept
  116. Shot 115, 13:29 to 13:38, 1 of 1 keyframes kept
  117. Shot 116, 13:38 to 13:40, 1 of 1 keyframes kept
  118. Shot 117, 13:40 to 13:43, 1 of 1 keyframes kept
  119. Shot 118, 13:43 to 13:48, 1 of 1 keyframes kept
  120. Shot 119, 13:48 to 13:52, 1 of 1 keyframes kept
  121. Shot 120, 13:52 to 13:55, 1 of 1 keyframes kept
  122. Shot 121, 13:55 to 14:15, 1 of 1 keyframes kept
  123. Shot 122, 14:15 to 14:30, 1 of 1 keyframes kept
  124. Shot 123, 14:30 to 14:39, 1 of 1 keyframes kept
  125. Shot 124, 14:39 to 14:48, 1 of 1 keyframes kept
  126. Shot 125, 14:48 to 14:52, 1 of 1 keyframes kept
  127. Shot 126, 14:52 to 14:54, 1 of 1 keyframes kept
  128. Shot 127, 14:54 to 15:06, 1 of 1 keyframes kept
  129. Shot 128, 15:06 to 15:10, 1 of 1 keyframes kept
  130. Shot 129, 15:10 to 15:17, 1 of 1 keyframes kept
  131. Shot 130, 15:17 to 15:20, 1 of 1 keyframes kept
  132. Shot 131, 15:20 to 15:27, 1 of 1 keyframes kept
  133. Shot 132, 15:27 to 15:28, 1 of 1 keyframes kept
  134. Shot 133, 15:28 to 15:29, 1 of 1 keyframes kept
  135. Shot 134, 15:29 to 15:31, 1 of 1 keyframes kept
  136. Shot 135, 15:31 to 15:44, 1 of 1 keyframes kept
  137. Shot 136, 15:44 to 15:46, 0 of 1 keyframes kept
  138. Shot 137, 15:46 to 15:48, 1 of 1 keyframes kept
  139. Shot 138, 15:48 to 15:51, 1 of 1 keyframes kept
  140. Shot 139, 15:51 to 16:08, 1 of 1 keyframes kept
  141. Shot 140, 16:08 to 16:11, 1 of 1 keyframes kept
  142. Shot 141, 16:11 to 16:14, 1 of 1 keyframes kept
  143. Shot 142, 16:14 to 16:16, 1 of 1 keyframes kept
  144. Shot 143, 16:16 to 16:24, 1 of 1 keyframes kept
  145. Shot 144, 16:24 to 16:36, 1 of 1 keyframes kept
  146. Shot 145, 16:36 to 16:39, 1 of 1 keyframes kept
  147. Shot 146, 16:39 to 16:44, 0 of 1 keyframes kept
  148. Shot 147, 16:44 to 16:48, 1 of 1 keyframes kept
  149. Shot 148, 16:48 to 17:00, 0 of 1 keyframes kept
  150. Shot 149, 17:00 to 17:12, 1 of 1 keyframes kept
  151. Shot 150, 17:12 to 17:31, 1 of 1 keyframes kept
  152. Shot 151, 17:31 to 17:46, 1 of 1 keyframes kept
  153. Shot 152, 17:46 to 17:49, 1 of 1 keyframes kept
  154. Shot 153, 17:49 to 17:50, 1 of 1 keyframes kept
  155. Shot 154, 17:50 to 18:10, 1 of 1 keyframes kept
  156. Shot 155, 18:10 to 18:15, 1 of 1 keyframes kept
  157. Shot 156, 18:15 to 18:17, 1 of 1 keyframes kept
  158. Shot 157, 18:17 to 18:27, 1 of 1 keyframes kept
  159. Shot 158, 18:27 to 18:31, 1 of 1 keyframes kept
  160. Shot 159, 18:31 to 18:36, 1 of 1 keyframes kept
  161. Shot 160, 18:36 to 18:41, 1 of 1 keyframes kept
  162. Shot 161, 18:41 to 18:54, 1 of 1 keyframes kept
  163. Shot 162, 18:54 to 19:00, 1 of 1 keyframes kept
  164. Shot 163, 19:00 to 19:17, 0 of 1 keyframes kept

164 shot(s).

keyframes kept every frame deduplicated

What was stored

cues
230
whisperx 230
chunks
34
from 230 cues
keyframes
152
kept of 164 captured
frames with text
152
4,794 lines read
chapters
11
from the source metadata
keyframe bytes
23.4 MB
word timings on 230 cues

Provenance

Each pipeline stage, its state and the model that produced it
stage state model started took
fetch done — 2026-08-09 23:13 0s
stt done — 2026-08-09 13:42 29s
chunk done — 2026-08-09 13:43 0s
text_embed done — 2026-08-10 19:41 1s
keyframe done — 2026-08-09 13:43 4m 03s
ocr done — 2026-08-09 13:47 1m 24s
frame_embed done — 2026-08-10 19:41 25s

Frames, and what the machine read

  • 0:02 #0 done2 line(s)

    shot 0·sharpness 455.1

    1. AlEngineer0.96
    2. World's Fair0.97

  • 0:03 #1 done2 line(s)

    shot 1·sharpness 656.7

    1. AlEngineer0.96
    2. World's Fair0.99

  • 0:10 #2 done24 line(s)

    shot 2·sharpness 2734.7

    1. LAB & PLATINUM SPONSORS0.99
    2. Amazon AGI Lab0.98
    3. ANTHROP\C0.99
    4. Google DeepMind1.00
    5. MINIMAX0.95
    6. OpenAI0.92
    7. Akamai1.00
    8. arize1.00
    9. aws1.00
    10. Braintrust bright data0.97
    11. B1.00
    12. Browserbase1.00
    13. docker1.00
    14. :neo4j0.92
    15. ORACLE1.00
    16. PayPal1.00
    17. qodo1.00
    18. reducto1.00
    19. Sonar1.00
    20. Makers of1.00
    21. togetherai1.00
    22. Unblocked1.00
    23. WorkOS1.00
    24. SonarQube1.00

  • 0:13 #3 done3 line(s)

    shot 3·sharpness 334.8

    1. AlEngineer0.96
    2. World's Fair1.00
    3. AIE1.00

  • 0:14 #4 done109 line(s)

    shot 4·sharpness 3077.1

    1. World's Fair0.92
    2. ORACLE1.00
    3. World's Fair0.96
    4. arize1.00
    5. World's Fair0.93
    6. Google DeepMind0.98
    7. World'sFai0.94
    8. :neo4]0.79
    9. World's Fair0.91
    10. Z.AI0.99
    11. World's Fair0.89
    12. bright data0.97
    13. World'sFain0.94
    14. erbose0.92
    15. World'sFa0.95
    16. extend1.00
    17. World'sFai1.00
    18. vast.ai1.00
    19. World's Fal0.90
    20. Ref.1.00
    21. RedHat1.00
    22. mezmo*0.89
    23. World'sFal0.93
    24. stigg1.00
    25. World'sFa1.00
    26. World'sFal0.94
    27. RELAI0.93
    28. Modal1.00
    29. THEVELOCITYROOM0.97
    30. fiddler1.00
    31. ZERO0.98
    32. comet1.00
    33. authe1.00
    34. SOIOIO0.92
    35. PRIOR1.00
    36. World's Fair0.99
    37. DigitalOcean1.00
    38. Vence0.94
    39. POSTHAN0.96
    40. Vorld'sF0.98
    41. World'sFa1.00
    42. Modular1.00
    43. Norld's Fal0.90
    44. MERGE1.00
    45. AUTOMATTIC1.00
    46. World's Fai0.95
    47. Zed1.00
    48. Keycard1.00
    49. Meticulous1.00
    50. Daytona1.00
    51. GRAVITEE1.00
    52. World'sFai0.99
    53. Llamalndex0.99
    54. World's Fai0.93
    55. INNGEST1.00
    56. World'sFair0.99
    57. BAND1.00
    58. World's Fai0.95
    59. Cleric0.99
    60. World'sFa0.94
    61. À ATLASSIAN0.92
    62. FACTORY1.00
    63. baseten1.00
    64. Snorkel0.99
    65. Z.AI0.99
    66. World'sF1.00
    67. qodo0.99
    68. PayPal1.00
    69. Gradium1.00
    70. ANTHROPIC0.99
    71. Amazon AGI Lab0.99
    72. :neo4j0.97
    73. Google DeepMi1.00
    74. World'sFai0.95
    75. arize1.00
    76. reducto1.00
    77. Microsoft1.00
    78. OpenAl0.98
    79. WorkOS1.00
    80. Amazon AGI Lab1.00
    81. World'sFa1.00
    82. Microsoft1.00
    83. ORACLE1.00
    84. bright data0.99
    85. Google DeepMind1.00
    86. Microsoft1.00
    87. docker1.00
    88. Braintrust1.00
    89. OpenAl0.99
    90. Z.AI0.98
    91. ANTHROPIC1.00
    92. snyk1.00
    93. togetherai1.00
    94. Lightrun1.00
    95. Norld'sFa0.91
    96. DATADOG1.00
    97. Resolve.ai1.00
    98. Vorid'sFa0.91
    99. World'sFa0.96
    100. LanceDB1.00
    101. World'sFai0.95
    102. builderio0.93
    103. Ravenna1.00
    104. descupe0.95
    105. World's Fal0.90
    106. TOPK0.91
    107. cognee1.00
    108. eNCORD0.94
    109. rlds Fa0.85

  • 0:14 #5 done183 line(s)

    shot 5·sharpness 3097.1

    1. torid's Fair0.88
    2. ORACLE1.00
    3. World'sFai0.96
    4. arize1.00
    5. World'sFair0.99
    6. Google DeepMind1.00
    7. World's Fair0.93
    8. :neo4j0.93
    9. World's Fair0.98
    10. Z.AI0.99
    11. World's Fai0.91
    12. bright data1.00
    13. World'sFal0.94
    14. er compute co.0.97
    15. World's Fair0.97
    16. extend1.00
    17. World's Fair0.95
    18. vast.ai1.00
    19. World's Fair0.96
    20. Ref.1.00
    21. RedHat1.00
    22. World's Fai0.97
    23. mezmo^{0.85
    24. World'sFa0.99
    25. stigg1.00
    26. World's Fa0.98
    27. rid's Fair0.95
    28. RELA0.99
    29. World's Fal0.93
    30. VAPI1.00
    31. World's Fain0.95
    32. Modal1.00
    33. promptql0.98
    34. World'sFair0.96
    35. THE VELOCITYROOM0.93
    36. World's Fair0.92
    37. fiddler1.00
    38. World's Fair0.96
    39. Supercandu0.95
    40. aurreal0.80
    41. ZERO0.98
    42. forld's Fai0.82
    43. comet1.00
    44. World's Fal0.96
    45. authe0.99
    46. World'sFa0.94
    47. fid'sFair0.91
    48. SOIO.IO0.90
    49. World'sFai0.97
    50. granica1.00
    51. World's Fair0.95
    52. dash00.99
    53. Vorld's Fair0.95
    54. PRIOR1.00
    55. World1.00
    56. 'sFair1.00
    57. jitalOcean0.93
    58. World'sFai0.99
    59. Vence0.97
    60. World's Fair0.94
    61. POSTMAN0.99
    62. World's Fair0.93
    63. Composio1.00
    64. World's Fa0.95
    65. 1d'sFai0.95
    66. Modular1.00
    67. World's Fal0.91
    68. MERGE1.00
    69. World'sFair0.99
    70. AUTOMATTIC1.00
    71. World's Fair0.97
    72. Buildk1.00
    73. gabyteDB0.98
    74. World'sFai1.00
    75. Zed1.00
    76. World'sFa0.91
    77. Keycard1.00
    78. World'sFa1.00
    79. Meticulous1.00
    80. World's Fa0.94
    81. 1d's Fair0.90
    82. Daytona1.00
    83. World'sFair1.00
    84. twilio0.95
    85. GRAVITEE1.00
    86. World's Fal0.88
    87. PlanetSco0.93
    88. Temporal1.00
    89. World's Fair0.91
    90. Llamalndex1.00
    91. World's Fai0.96
    92. OO INNGEST0.82
    93. World'sFair0.96
    94. BAND1.00
    95. World'sFair0.99
    96. Cleric1.00
    97. World's Fai0.96
    98. À ATLASSIAN0.95
    99. World's Fa0.92
    100. FACTORY1.00
    101. World's Fa0.91
    102. 1d'sFal0.85
    103. baseten1.00
    104. World'sFai0.95
    105. Snorkel0.97
    106. World'sFa1.00
    107. Z.AI0.99
    108. World'sFa1.00
    109. IDER0.86
    110. qodo0.98
    111. PayPal1.00
    112. World'sFa1.00
    113. Gradium1.00
    114. World'sFai0.94
    115. ANTHROP1.00
    116. in AGI Lab0.95
    117. World'sFa0.98
    118. Browserbase1.00
    119. World'sFai0.96
    120. :neo4j0.94
    121. World's Fair0.92
    122. Google DeepMing0.95
    123. World'sFa1.00
    124. arize1.00
    125. World'sFa0.98
    126. reducto1.00
    127. Microsoft1.00
    128. World's Fa0.95
    129. id'sFair0.78
    130. OpenAl0.98
    131. World'sFai0.96
    132. WorkOS0.99
    133. Amazon AGI Lab1.00
    134. World'sF0.99
    135. Microsoft0.98
    136. World'sFai0.96
    137. ORACLE1.00
    138. World'sFa1.00
    139. bright data0.99
    140. World'sFa1.00
    141. Google Deeph0.95
    142. Microsoft0.99
    143. World'sFa0.93
    144. docker1.00
    145. Braintrust1.00
    146. OpenAl1.00
    147. MINIMAX1.00
    148. Vorld'sFa0.96
    149. Z.AI0.98
    150. World's Fal0.94
    151. ANTHROPIC1.00
    152. World'sFa1.00
    153. 1d'sFa0.98
    154. snyk1.00
    155. togetherai1.00
    156. World'sFa0.99
    157. greptil0.96
    158. 'Lightrun0.99
    159. World'sFal0.95
    160. CLOUD0.98
    161. DATADOG1.00
    162. World's Fa0.96
    163. Resolve.ai1.00
    164. World's Fa0.92
    165. rid'sFal0.93
    166. 680.84
    167. LanceDB1.00
    168. World'sFa1.00
    169. builderio0.96
    170. World'sFai0.99
    171. Ravenna1.00
    172. World'sFai0.97
    173. CopilotKit0.99
    174. ScHpe0.60
    175. World's Fair0.96
    176. TOP1.00
    177. World'sFai0.94
    178. cognee1.00
    179. World'sFa1.00
    180. eNCORD0.96
    181. World'sFai0.95
    182. World's F0.86
    183. 100●0.54

  • 0:18 #6 done6 line(s)

    shot 6·sharpness 634.6

    1. DEX HORTHY0.99
    2. CO-FOUNDER AND CEO1.00
    3. humanlayer1.00
    4. AIE1.00
    5. World's Fair0.99
    6. Microsoft0.99

  • 0:19 #7 skipped

    shot 7·duplicate of #6

  • 0:20 #8 done6 line(s)

    shot 8·sharpness 651.5

    1. DEX HORTHY0.97
    2. CO-FOUNDER AND CEO1.00
    3. ≡humanlayer0.99
    4. AIE1.00
    5. World's Fair0.98
    6. Microsoft1.00

  • 0:28 #9 done50 line(s)

    shot 9·sharpness 1245.6

    1. World'sFair1.00
    2. Ref.1.00
    3. World's Fai0.93
    4. RedHat1.00
    5. Worl0.96
    6. Modal1.00
    7. World'sFair0.97
    8. promptql1.00
    9. World'sFair0.96
    10. THEVELO1.00
    11. AIEng0.92
    12. jineer0.99
    13. Norld1.00
    14. 's Fair0.96
    15. DEX HORTHY1.00
    16. World'sFair0.98
    17. BAND1.00
    18. World'sFair1.00
    19. Cleric1.00
    20. Worl0.99
    21. Z.AI0.97
    22. World'sFair1.00
    23. CO-FOUNDER AND CEO0.99
    24. qodo0.98
    25. World'sFai0.98
    26. Pa0.82
    27. World'sFair1.00
    28. Google DeepMind1.00
    29. umanlayer1.00
    30. World's Fair0.95
    31. arize1.00
    32. Worl0.92
    33. Amazon AGI Lab1.00
    34. World'sFair0.99
    35. Microsoft1.00
    36. World'sFa0.99
    37. OR1.00
    38. World'sFair1.00
    39. OpenAl0.94
    40. World'sFair0.99
    41. MINIMAX1.00
    42. Wort0.94
    43. togetherai0.98
    44. World'sFai0.99
    45. World1.00
    46. Wort0.92
    47. Ra1.00
    48. World's Fal0.91
    49. Wor1.00
    50. Microsoft1.00

  • 0:31 #10 done189 line(s)

    shot 10·sharpness 3443.3

    1. ute co.0.96
    2. World's Fair0.97
    3. extend1.00
    4. World'sFair0.94
    5. vast.ai1.00
    6. World's Fair0.98
    7. Ref.1.00
    8. World's Fair0.96
    9. RedHat1.00
    10. World's Fair0.97
    11. mezmo*0.97
    12. World's Fair0.96
    13. stigg1.00
    14. World0.99
    15. RELAI0.99
    16. World's Fair0.98
    17. VAPI0.99
    18. World's Fair0.97
    19. Modal1.00
    20. World's Fair0.93
    21. promptql1.00
    22. World's Fair0.95
    23. THE VELOCITY ROOM0.97
    24. World's Fair0.95
    25. fiddler1.00
    26. World's Fair0.95
    27. Superce0.94
    28. DB0.99
    29. World's Fair0.94
    30. ZERO0.98
    31. World's Fair0.91
    32. AlEngineer0.97
    33. comet1.00
    34. World's Fair0.93
    35. authe0.95
    36. World1.00
    37. SOIO.IO0.81
    38. World's Fair0.99
    39. granica1.00
    40. World's Fair0.93
    41. dash00.99
    42. World's Fair0.97
    43. PRI0.99
    44. World's Fair1.00
    45. cean1.00
    46. World's Fair0.96
    47. Vence1.00
    48. World's Fair0.96
    49. POSTMAN0.99
    50. World's Fair0.92
    51. Composio1.00
    52. World1.00
    53. Modular0.94
    54. World's Fair0.97
    55. MERGE1.00
    56. World's Fair0.99
    57. AUTOMATTIC1.00
    58. World's Fair0.98
    59. Bui0.94
    60. eDB1.00
    61. World's Fair0.98
    62. Zed0.97
    63. World's Fair0.97
    64. Keycard1.00
    65. World's Fair0.95
    66. Meticulous1.00
    67. World0.99
    68. air0.92
    69. Daytona1.00
    70. World's Fair0.98
    71. twilio0.84
    72. World's Fair0.99
    73. G GRAVITEE0.89
    74. World's Fair0.95
    75. Plane0.99
    76. oral0.97
    77. World's Fair0.98
    78. I Llamalndex0.97
    79. World's Fair0.96
    80. INNGEST0.94
    81. World's Fair0.95
    82. ©BAND0.91
    83. World's Fair0.98
    84. Cleric1.00
    85. World's Fair0.98
    86. À ATLASSIAN0.96
    87. World's Fair0.93
    88. FACTORY1.00
    89. World1.00
    90. a0.76
    91. baseten1.00
    92. World's Fair0.98
    93. Snorkel0.99
    94. World's Fair0.93
    95. Z.AI0.98
    96. World's Fair0.99
    97. qodo0.99
    98. World's Fair0.95
    99. PayPal1.00
    100. World's Fair0.97
    101. Gradium1.00
    102. World's Fair0.97
    103. ANTHR0.98
    104. I Lab0.90
    105. World's Fair0.95
    106. Browserbase1.00
    107. World's Fair0.99
    108. :neo4j0.98
    109. World's Fair0.98
    110. Google DeepMind1.00
    111. World's Fair0.98
    112. arize1.00
    113. World's Fair0.99
    114. reducto1.00
    115. World's Fair0.99
    116. Microsoft0.93
    117. World1.00
    118. OpenAI0.92
    119. World's Fair0.97
    120. WorkOS0.89
    121. World's Fair0.99
    122. zon AGI Lab0.99
    123. World's Fair0.96
    124. Microsoft1.00
    125. World's Fair0.98
    126. ORACLE1.00
    127. World's Fair0.95
    128. bright data0.97
    129. World'sFair0.99
    130. Google D0.97
    131. oft1.00
    132. World's Fair0.94
    133. docker1.00
    134. World's Fair0.97
    135. Braintrust1.00
    136. orld's Fair0.97
    137. OpenAl0.93
    138. World's Fair0.98
    139. MINIMAX0.96
    140. World'sFair1.00
    141. Z.AI0.99
    142. World's Fair0.99
    143. ANTHROPIC0.97
    144. World1.00
    145. snyk1.00
    146. World's Fair0.96
    147. aws0.99
    148. World's Fair0.99
    149. ogetherai0.95
    150. World's Fair0.97
    151. Akamai0.99
    152. World's Fair0.96
    153. Unblocked0.99
    154. run1.00
    155. World's Fair0.97
    156. CLOUDPLARE0.96
    157. orld's Fair0.97
    158. DATADOG1.00
    159. Worid's Fair0.95
    160. Resolve.ai1.00
    161. World's Fair0.95
    162. World's Fa0.94
    163. Op0.98
    164. World0.99
    165. LanceDB1.00
    166. World's Fair0.94
    167. builder.io0.90
    168. World's Fair0.99
    169. Ravenna1.00
    170. World's Fair0.95
    171. CopilotKit1.00
    172. VERIS1.00
    173. Wor0.98
    174. World's Fair0.91
    175. e1.00
    176. Mirosoft0.96
    177. t1.00
    178. World's Fair0.99
    179. TOPK1.00
    180. Vorld's Fair0.97
    181. cognee1.00
    182. Ward's Fair0.89
    183. eNCORD0.95
    184. World's Fair0.98
    185. rid'sFa0.88
    186. World1.00
    187. ase0.94
    188. 90.57
    189. OG0.53

  • 0:33 #11 done11 line(s)

    shot 11·sharpness 414.4

    1. woridsFair0.96
    2. oogie Deepriind0.86
    3. AlEngineer0.98
    4. Amazon AGI Lab1.00
    5. Vorld'sFair1.00
    6. AIEngineer0.95
    7. OpenAl0.92
    8. World'sFai1.00
    9. AlEngineer0.97
    10. together1.00
    11. World'sFair1.00

  • 0:36 #12 done9 line(s)

    shot 12·sharpness 514.9

    1. AlEngineer0.98
    2. World's Fair1.00
    3. Microsoft1.00
    4. ineer1.00
    5. OpenAl0.91
    6. Fair1.00
    7. AlEngineer0.99
    8. AL mai0.82
    9. World'sF1.00

  • 0:41 #13 done6 line(s)

    shot 13·sharpness 503.5

    1. AlEngineer0.96
    2. AI0.73
    3. rld's Fair0.99
    4. Am: AGI Lab0.94
    5. Wor1.00
    6. Op0.99

  • 0:54 #14 done

    shot 14·sharpness 2283.9

  • 1:01 #15 done

    shot 15·sharpness 4748.8

  • 1:02 #16 done

    shot 16·sharpness 4995.5

  • 1:08 #17 done

    shot 17·sharpness 4289.7

  • 1:16 #18 done

    shot 18·sharpness 2668.9

  • 1:18 #19 done

    shot 19·sharpness 4224.4

  • 1:24 #20 done

    shot 20·sharpness 2680.5

  • 1:30 #21 done

    shot 21·sharpness 2959.8

  • 1:32 #22 done

    shot 22·sharpness 3483.0

  • 1:35 #23 done

    shot 23·sharpness 3807.0

The page's on-screen-text budget of 600 lines is spent, so the last cards in this grid list fewer lines than they hold. Narrow the page with ?frames= to read them.

Transcript

230 cues· 3,821 words· 20,425 chars

  1. 0:25 Guys give it up for all the great speakers today so far.
  2. 0:36 Alright this is a harness engineering is not enough and why software factories fail.
  3. 0:42 And we're going to click maybe.
  4. 0:47 Oh that's way too many slides hold on guys.
  5. 0:50 Okay, so we're all racing to put AI coding into production, and there's been lots been said about loop engineering, and we should probably write more loops, and yeah, I don't know, I guess we're doing loops now.
  6. 1:06 StrongDM built a lights-off software factory where nobody even reads the code, and the prevailing narrative is we should just spend more tokens, you are the bottleneck,
  7. 1:17 The models are good enough, code is free.
  8. 1:20 Just ship more stuff.
  9. 1:22 But at the same time, we are starting to see the cracks.
  10. 1:26 Our friend Mario at AI Engineer Europe begged us to slow down because companies that should not be having outages because of coding agents are having outages due to coding agent mishaps.
  11. 1:37 Codebases are falling apart faster than they ever have before.
  12. 1:41 Our friends at Pharos AI actually even did a report since we all picked up all these AI coding tools in January, maybe February.
  13. 1:49 Pull request code review quality is way down.
  14. 1:51 We're having more comments, longer comments, and tons of PRs being merged without any review at all.
  15. 1:57 Incidents are way up, bugs per developer are way up, and many people will tell you that you're holding it wrong.
  16. 2:04 That's the only reason, you're not.
  17. 2:07 Well, maybe you are, but that's not the point.
  18. 2:12 I've spoken a lot about how to hold it better when it comes to working with AI, probably a million views on YouTube at this point across a bunch of different talks.
  19. 2:20 And the basic thing is like as engineers, we've been told that if token maxing isn't working, then it's a skill issue.
  20. 2:27 You just need to spend more tokens.
  21. 2:29 Let go of reading the code.
  22. 2:31 That with enough harness engineering,
  23. 2:33 if we maybe sprinkle some magic words, adversarial review on enough of our PR bots, that we can get the best of both worlds.
  24. 2:42 10 to 100x faster, high quality, and nobody has to do that thing we all hate called code review.
  25. 2:49 I'm here to convince you today that this is in fact not a skill issue.
  26. 2:53 That no amount of harness engineering or loops maxing can solve what is fundamentally a model training issue.
  27. 3:00 That's why we say the harness is not enough.
  28. 3:03 And to understand this, we kind of have to grapple and dig into how coding models are trained.
  29. 3:08 I'm gonna talk about what I think the shortcomings are with some of the current benchmarks and what better ones might look like.
  30. 3:13 And we'll talk about how to move faster safely in the meantime.
  31. 3:17 It's gonna sound like a rant, but there is hope here.
  32. 3:19 I'm gonna talk about our journey and a bunch of the landmines we've hit building in this world, a bunch of exciting new techniques that we've been working with a lot of our users and customers to develop, and I think how we all as a community get to the next chapter of agentic engineering after whatever this thing that we're in.
  33. 3:36 So we use a lot of words here.
  34. 3:37 I'm going to zoom out a little bit.
  35. 3:38 I want to give you kind of like a brief history of the software factory.
  36. 3:42 And actually, I just learned this last week.
  37. 3:44 The term software factory was defined at a NATO conference in 1968.
  38. 3:47 We're going to start around 2022, right before AI started coming around.
  39. 3:53 And basically, in a typical 2022 software factory,
  40. 3:57 you will have some people building stuff.
  41. 3:59 You'll have engineers, you'll have PMs, maybe you have some leadership team that is driving the vision here, and they all decide that stuff needs to get done.
  42. 4:07 So you put it in a tracker, a linear, a JIRA, a beads, some state machine that tracks what needs to be done.
  43. 4:13 And then someone goes and grabs something off of there, and they build the thing.
  44. 4:17 And there may be some automated testing in that process, maybe some manual testing in that process.
  45. 4:21 At a certain point, we make this pull request thing.
  46. 4:24 It says, okay, cool, we've got to run a bunch of checks, automated stuff.
  47. 4:27 A human's going to review the change and review the code.
  48. 4:29 And perhaps we might even have a human pull it down and test it somehow.
  49. 4:34 And if anything goes wrong here, we loop back to someone builds the thing.
  50. 4:38 And eventually we're ready for prod.

Chapters

  1. 0:00 The narrative: you are the bottleneck, just ship more
  2. 1:28 The cracks: outages and falling PR review quality
  3. 2:20 The thesis: the harness is not enough
  4. 3:36 A brief history of the software factory
  5. 5:52 The agentic factory and turning the lights off
  6. 7:30 Why it fails: the July 2025 lights-off experiment
  7. 8:56 Models cannot maintain codebase quality
  8. 10:12 Why Claude Code won and how coding models are trained
  9. 13:18 Verifying maintainability and better benchmarks
  10. 14:58 Turning the lights back on: plan up front
  11. 17:16 Too many bad PRs, and closing advice

Open at this second