vidtheque.

read-only demo

Videos pWXUkLP9uWM

First Steps Toward Automated AI Research — Richard Socher, CEO Recursive AI

index_state ready data_status ok

AI Engineer· published 2026-07-30· 0:20:23· en-US· indexed 2026-08-10 19:38

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 0:20, 1 of 1 keyframes kept
  5. Shot 4, 0:20 to 0:30, 1 of 1 keyframes kept
  6. Shot 5, 0:30 to 1:12, 1 of 1 keyframes kept
  7. Shot 6, 1:12 to 1:44, 1 of 1 keyframes kept
  8. Shot 7, 1:44 to 2:16, 0 of 1 keyframes kept
  9. Shot 8, 2:16 to 2:20, 1 of 1 keyframes kept
  10. Shot 9, 2:20 to 2:26, 1 of 1 keyframes kept
  11. Shot 10, 2:26 to 2:34, 0 of 1 keyframes kept
  12. Shot 11, 2:34 to 2:43, 1 of 1 keyframes kept
  13. Shot 12, 2:43 to 2:55, 1 of 1 keyframes kept
  14. Shot 13, 2:55 to 3:23, 0 of 1 keyframes kept
  15. Shot 14, 3:23 to 3:50, 1 of 1 keyframes kept
  16. Shot 15, 3:50 to 4:18, 0 of 1 keyframes kept
  17. Shot 16, 4:18 to 4:51, 1 of 1 keyframes kept
  18. Shot 17, 4:51 to 5:22, 1 of 1 keyframes kept
  19. Shot 18, 5:22 to 5:51, 1 of 1 keyframes kept
  20. Shot 19, 5:51 to 6:26, 1 of 1 keyframes kept
  21. Shot 20, 6:26 to 6:39, 0 of 1 keyframes kept
  22. Shot 21, 6:39 to 7:07, 1 of 1 keyframes kept
  23. Shot 22, 7:07 to 7:36, 1 of 1 keyframes kept
  24. Shot 23, 7:36 to 7:57, 1 of 1 keyframes kept
  25. Shot 24, 7:57 to 8:25, 1 of 1 keyframes kept
  26. Shot 25, 8:25 to 9:06, 1 of 1 keyframes kept
  27. Shot 26, 9:06 to 9:24, 1 of 1 keyframes kept
  28. Shot 27, 9:24 to 9:53, 1 of 1 keyframes kept
  29. Shot 28, 9:53 to 10:22, 0 of 1 keyframes kept
  30. Shot 29, 10:22 to 10:24, 1 of 1 keyframes kept
  31. Shot 30, 10:24 to 10:39, 1 of 1 keyframes kept
  32. Shot 31, 10:39 to 11:05, 1 of 1 keyframes kept
  33. Shot 32, 11:05 to 11:32, 1 of 1 keyframes kept
  34. Shot 33, 11:32 to 11:53, 1 of 1 keyframes kept
  35. Shot 34, 11:53 to 12:26, 1 of 1 keyframes kept
  36. Shot 35, 12:26 to 13:00, 1 of 1 keyframes kept
  37. Shot 36, 13:00 to 13:43, 1 of 1 keyframes kept
  38. Shot 37, 13:43 to 13:55, 1 of 1 keyframes kept
  39. Shot 38, 13:55 to 14:21, 1 of 1 keyframes kept
  40. Shot 39, 14:21 to 14:46, 1 of 1 keyframes kept
  41. Shot 40, 14:46 to 15:11, 1 of 1 keyframes kept
  42. Shot 41, 15:11 to 15:37, 0 of 1 keyframes kept
  43. Shot 42, 15:37 to 16:10, 1 of 1 keyframes kept
  44. Shot 43, 16:10 to 16:43, 1 of 1 keyframes kept
  45. Shot 44, 16:43 to 17:15, 1 of 1 keyframes kept
  46. Shot 45, 17:15 to 17:44, 1 of 1 keyframes kept
  47. Shot 46, 17:44 to 18:14, 0 of 1 keyframes kept
  48. Shot 47, 18:14 to 18:43, 0 of 1 keyframes kept
  49. Shot 48, 18:43 to 19:08, 1 of 1 keyframes kept
  50. Shot 49, 19:08 to 19:33, 1 of 1 keyframes kept
  51. Shot 50, 19:33 to 19:58, 1 of 1 keyframes kept
  52. Shot 51, 19:58 to 20:06, 1 of 1 keyframes kept
  53. Shot 52, 20:06 to 20:23, 0 of 1 keyframes kept

53 shot(s).

keyframes kept every frame deduplicated

What was stored

cues
149
whisperx 149
chunks
36
from 149 cues
keyframes
43
kept of 53 captured
frames with text
43
1,456 lines read
chapters
12
from the source metadata
keyframe bytes
8.9 MB
word timings on 149 cues

Provenance

Each pipeline stage, its state and the model that produced it
stage state model started took
fetch done — 2026-08-09 21:52 0s
stt done — 2026-08-09 03:03 27s
chunk done — 2026-08-09 03:03 0s
text_embed done — 2026-08-10 19:37 0s
keyframe done — 2026-08-09 03:03 2m 50s
ocr done — 2026-08-09 03:06 22s
frame_embed done — 2026-08-10 19:37 8s

Frames, and what the machine read

  • 0:02 #0 done2 line(s)

    shot 0·sharpness 456.7

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

  • 0:03 #1 done2 line(s)

    shot 1·sharpness 663.3

    1. AlEngineer0.95
    2. World's Fair0.99

  • 0:10 #2 done25 line(s)

    shot 2·sharpness 2742.9

    1. LAB & PLATINUM SPONSORS0.99
    2. Amazon AGI Lab0.97
    3. ANTHROP\C1.00
    4. Google DeepMind1.00
    5. MINIMAX0.98
    6. OpenAl0.89
    7. Akamai1.00
    8. arize1.00
    9. aws1.00
    10. Braintrust1.00
    11. bright data0.99
    12. B1.00
    13. Browserbase1.00
    14. docker1.00
    15. :neo4j0.90
    16. ORACLE1.00
    17. PayPal1.00
    18. qodo1.00
    19. reducto1.00
    20. Sonar1.00
    21. Makers of1.00
    22. togetherai1.00
    23. Unblocked1.00
    24. WorkOS1.00
    25. SonarQube1.00

  • 0:13 #3 done89 line(s)

    shot 3·sharpness 4117.1

    1. World's Fair0.98
    2. Modular1.00
    3. World's Fair0.98
    4. MERGE1.00
    5. voriuSr0.76
    6. yugabyteDB1.00
    7. World's Fair0.98
    8. Zed0.98
    9. World's Fair0.99
    10. World's Fair0.96
    11. Daytona1.00
    12. World's Fair0.96
    13. twilio0.94
    14. Temporal1.00
    15. World's Fair0.97
    16. Llamalndex1.00
    17. World's Fair0.98
    18. INNGEST1.00
    19. World's Fair0.98
    20. BAND1.00
    21. World'sFair1.00
    22. World'sFair1.00
    23. baseten1.00
    24. World's Fair0.97
    25. Snorkel0.98
    26. World's Fair0.98
    27. Z.AI0.99
    28. World's Fair0.97
    29. qodo0.99
    30. Wo1.00
    31. Amazon AGI Lab0.99
    32. World's Fair0.96
    33. Browserbase1.00
    34. World's Fair0.98
    35. :neo4j0.96
    36. World's Fair0.96
    37. Google DeepMind1.00
    38. World'sFair1.00
    39. World'sFair1.00
    40. OpenAl0.93
    41. World's Fair0.96
    42. WorkOS0.99
    43. World's Fair0.99
    44. Amazon AGI Lab1.00
    45. World's Fair0.99
    46. Microsoft1.00
    47. Wo1.00
    48. Microsoft1.00
    49. World'sFai0.96
    50. docker1.00
    51. World's Fair0.97
    52. Braintrust1.00
    53. World's Fair0.97
    54. OpenAl0.94
    55. World's Fair0.98
    56. 小0.82
    57. World'sFair1.00
    58. snyk1.00
    59. World's Fair0.97
    60. aws1.00
    61. World's Fair0.97
    62. togetherai1.00
    63. World's Fair0.99
    64. Akamai1.00
    65. Wo1.00
    66. Lightrun1.00
    67. World's Fair0.98
    68. World'sFair0.99
    69. DATADOG1.00
    70. World's Fair0.97
    71. R1.00
    72. CLOUDFLARE1.00
    73. World'sFair1.00
    74. World's Fair0.97
    75. ir0.73
    76. LanceDB1.00
    77. World's Fair0.96
    78. builder.io1.00
    79. World's Fair0.97
    80. Ravenna1.00
    81. Wo1.00
    82. descope0.99
    83. PRESENTED BY1.00
    84. rosoft1.00
    85. World's Fair0.96
    86. TOPK1.00
    87. World's Fair0.98
    88. cognee1.00
    89. World's Fair0.98

  • 0:21 #4 done65 line(s)

    shot 4·sharpness 2752.4

    1. World's Fair0.99
    2. Z.AI0.99
    3. World's Fair0.98
    4. qodo0.99
    5. World's Fair0.98
    6. PayPal1.00
    7. World1.00
    8. AlEng0.94
    9. :neo4j0.96
    10. World's Fair0.98
    11. AlEngineer0.94
    12. Google DeepMind1.00
    13. World's Fair0.94
    14. AlEngineer0.98
    15. arize1.00
    16. World'sFair1.00
    17. red1.00
    18. AlEngineer0.99
    19. AlEngineer1.00
    20. AlEngineer1.00
    21. AlEng0.99
    22. World's Fair0.98
    23. Amazon AGI Lab0.99
    24. World'sFair1.00
    25. Microsoft1.00
    26. World's Fair0.97
    27. ORACLE1.00
    28. World1.00
    29. Braintrust1.00
    30. World'sFair1.00
    31. OpenAI0.92
    32. rld's Fair0.95
    33. MINIMAX1.00
    34. World'sFair1.00
    35. Z.0.99
    36. World's Fair0.97
    37. together.ai0.92
    38. World'sFa1.00
    39. AlEngineer0.99
    40. kamai1.00
    41. World's Fair0.96
    42. Unblocked1.00
    43. World1.00
    44. AlEng0.99
    45. AlEngineer1.00
    46. World'sFair1.00
    47. 福0.85
    48. DATADOG1.00
    49. orld'sFair0.99
    50. Resolve.ai1.00
    51. World'sFair1.00
    52. CLOUDFLARE1.00
    53. World's Fair0.97
    54. builder.io0.97
    55. World's Fai0.96
    56. AlEngineer0.99
    57. :Ravenna0.95
    58. World'sFair1.00
    59. CopilotKit1.00
    60. TOPK1.00
    61. World'sFair1.00
    62. 小cognee0.92
    63. orld's Fair0.96
    64. enc0.80
    65. World'sFair1.00

  • 0:39 #5 done19 line(s)

    shot 5·sharpness 1990.3

    1. AlEngineer0.97
    2. recursive1.00
    3. World'sFair1.00
    4. The Eureka Machine:0.98
    5. Evolution and Automated Research1.00
    6. AlEngine0.95
    7. WorkOS0.94
    8. World's1.00
    9. Richard Socher1.00
    10. Recursive, You.com, and AIX Ventures0.99
    11. July1 20260.95
    12. AlEngineer0.99
    13. orld's Fai0.95
    14. Brain1.00
    15. AlEnginer0.97
    16. Engineering the future of Al1.00
    17. aws0.99
    18. rld's0.98
    19. AlEngineer0.99

  • 1:19 #6 done12 line(s)

    shot 6·sharpness 1685.8

    1. AlEngineer0.99
    2. World'sFair1.00
    3. Open-Ended Evolution0.99
    4. From Biology to Science, Technology:and Al0.97
    5. AlEngineer0.99
    6. World'sFair1.00
    7. Amazc1.00
    8. #Braintr0.96
    9. Wor1.00
    10. AlEngineer0.98
    11. Engineering the future of Al1.00
    12. World'sF0.98

  • 1:48 #7 skipped

    shot 7·duplicate of #6

  • 2:17 #8 done41 line(s)

    shot 8·sharpness 6048.0

    1. Biological evolution1.00
    2. recursive1.00
    3. AlEngineer0.94
    4. World's Fair0.96
    5. Echinoderms1.00
    6. Fish1.00
    7. Amphibians1.00
    8. Reptiles0.97
    9. Birds1.00
    10. Eukaryotes1.00
    11. Mammals1.00
    12. Archaea1.00
    13. Bacteria1.00
    14. Oceans Rust1.00
    15. AlEngineer1.00
    16. Earth Birth0.98
    17. rld's Fair0.99
    18. nazon A(0.95
    19. Today0.99
    20. 201 2520.97
    21. 370 4440.94
    22. 541 700 10000.93
    23. 20000.99
    24. 30001.00
    25. Millions of Years Ago1.00
    26. 40000.87
    27. 30000.99
    28. 20000.93
    29. 1000 700 541 4440.94
    30. 2521.00
    31. 2011.00
    32. 661.00
    33. Today1.00
    34. -AlEngineer0.92
    35. raintrust1.00
    36. World's0.92
    37. Altr0.94
    38. Engineering the future of Al0.99
    39. rld's0.99
    40. togeth1.00
    41. AlEngineer1.00

  • 2:25 #9 done41 line(s)

    shot 9·sharpness 6268.1

    1. Biological evolution1.00
    2. recursive1.00
    3. AlEngineer0.99
    4. World'sFair1.00
    5. Echinoderms0.99
    6. Protostomes1.00
    7. Fish1.00
    8. Amphibians1.00
    9. Reptiles1.00
    10. Birds1.00
    11. Eukaryotes1.00
    12. Mammals1.00
    13. Archaea1.00
    14. Bacteria1.00
    15. Oceans Rust1.00
    16. AlEngineer1.00
    17. Earth Birth0.98
    18. rkOS0.98
    19. World's Fair0.98
    20. Today1.00
    21. 661.00
    22. 541 700 10000.97
    23. 20000.98
    24. 30001.00
    25. Millions of Years Ago1.00
    26. 40000.89
    27. 30001.00
    28. 20001.00
    29. 1000 700 541 4440.95
    30. 3701.00
    31. 2521.00
    32. 2011.00
    33. 661.00
    34. Today1.00
    35. sFair0.97
    36. Braintrus1.00
    37. 'S0.55
    38. Engineer1.00
    39. d's Fair0.96
    40. Engineering the future of Al0.98
    41. neer0.90

  • 2:28 #10 skipped

    shot 10·duplicate of #9

  • 2:41 #11 done35 line(s)

    shot 11·sharpness 1883.1

    1. Biological evolution1.00
    2. recursive1.00
    3. AlEngineer0.99
    4. World'sFair1.00
    5. Echinoderms1.00
    6. Protostomes1.00
    7. Fish1.00
    8. Amphibians1.00
    9. Fungi1.00
    10. Reptiles0.94
    11. Plants1.00
    12. Birds1.00
    13. Eukaryotes1.00
    14. Mammals1.00
    15. Archaea1.00
    16. Bacteria1.00
    17. Oceans Rust0.99
    18. We arehere0.98
    19. AlEngineer0.99
    20. Earth Birth1.00
    21. WorkC0.88
    22. World's1.00
    23. 20000.97
    24. 30000.98
    25. Millions of Years Ago1.00
    26. 30000.91
    27. 20001.00
    28. 10000.93
    29. AlEngineer0.99
    30. rld's1.00
    31. Braint1.00
    32. AlEngineer0.97
    33. Engineering the future of Al0.98
    34. World'sI0.97
    35. AlEngine0.98

  • 2:54 #12 done43 line(s)

    shot 12·sharpness 2735.2

    1. Technologicalevolution1.00
    2. recursive1.00
    3. AlEngineer0.95
    4. World's Fair0.98
    5. Fig. 1 World Product, Data vs. Models0.98
    6. CES Combined Exp Model0.99
    7. Hyperbolic Model1.00
    8. —· Sum of Exp Model0.95
    9. +World Product Estimates0.98
    10. 10,000,000,000,0001.00
    11. 1,000,000,000,0001.00
    12. 100,000,000,0001.00
    13. 10,000,000,0001.00
    14. Industrial1.00
    15. Revolution1.00
    16. 1,000,000,0001.00
    17. 100,000,0001.00
    18. Farming1.00
    19. Enlightenment1.00
    20. 10,000,0001.00
    21. Hunting1.00
    22. 1,000,0001.00
    23. 100,0001.00
    24. AlEngine1.00
    25. 10,0001.00
    26. zon AGI Lal0.96
    27. World's1.00
    28. 10,000,0001.00
    29. 1,000,0000.99
    30. 100,0001.00
    31. 10,0001.00
    32. 1,0001.00
    33. 1001.00
    34. 101.00
    35. Hanson, R. (2020). Long-Term Growth As A Sequence of Exponential Modes.1.00
    36. orld's Fai0.97
    37. AlEngineer1.00
    38. Oper1.00
    39. AlEngine1.00
    40. Engineering the future of Al1.00
    41. togethe1.00
    42. Orld's0.91
    43. AlEngineer0.99

  • 3:14 #13 skipped

    shot 13·duplicate of #12

  • 3:29 #14 done71 line(s)

    shot 14·sharpness 3032.0

    1. Technologicalevolution → More people0.98
    2. recursive1.00
    3. AlEngineer0.99
    4. World's Fair0.99
    5. 6,0001.00
    6. Genome Project1.00
    7. 5,0001.00
    8. PCs1.00
    9. Pouli(n tons0.53
    10. Man on Moon0.99
    11. 4,0001.00
    12. High Speed Computers1.00
    13. Nuclear Energy0.98
    14. 3,0001.00
    15. Peak of Greece0.97
    16. Peak of Rome0.96
    17. Discovery of1.00
    18. New World1.00
    19. War on Malaria1.00
    20. Discovery of DNA1.00
    21. Penicillin0.99
    22. 2,0001.00
    23. Plague1.00
    24. Black1.00
    25. Invention of Airplane1.00
    26. Invention of Automobile1.00
    27. Invention of Telephone Electrification1.00
    28. 1,0001.00
    29. Beginning of 1st1.00
    30. Agricultural1.00
    31. Revolution1.00
    32. Beginning of1.00
    33. Pottery0.96
    34. Invention1.00
    35. of Plow0.97
    36. 1st Irrigation0.97
    37. Works1.00
    38. 1st Cities1.00
    39. Beginning of0.99
    40. Metallurgy1.00
    41. Beginning of1.00
    42. Writing1.00
    43. Beginning of1.00
    44. Mathematics1.00
    45. Beginning of 2nd Industrial Revolution1.00
    46. Beginning of 2nd Agricultural Revolution1.00
    47. Invention of Watt Engine1.00
    48. Beginning of Railroads1.00
    49. Germ Theory1.00
    50. -9,0001.00
    51. -6,0001.00
    52. -5,0001.00
    53. -4,0001.00
    54. -3,0001.00
    55. -2,0001.00
    56. -1,0001.00
    57. 00.97
    58. 1,0001.00
    59. 2,0001.00
    60. AlEngineer1.00
    61. orkOS0.89
    62. World's Fai0.98
    63. jineer0.96
    64. Fogel, R. W. (1999). Catching Up with the Economy. American Economic Review, 89(1), 1-21. https://doi.org/10.1257/aer.89.1.10.99
    65. I'sFa0.98
    66. Braintrus1.00
    67. VS0.86
    68. lorld's Fai0.95
    69. Engineer0.99
    70. Engineering the future of Al0.99
    71. jineer1.00

  • 3:59 #15 skipped

    shot 15·duplicate of #14

  • 4:41 #16 done21 line(s)

    shot 16·sharpness 2724.4

    1. Technological evolution → Growth0.97
    2. recursive1.00
    3. AlEngineer0.95
    4. World'sFair1.00
    5. "Theonly perpetual source of growthis technology.0.99
    6. Infact,technology - new knowledge, new tools, what theGreeks called techne- has always been the0.99
    7. main source of growth, and perhaps the only cause of growth, as technology made both population1.00
    8. growth and natural resource utilization possible."1.00
    9. AlEngineer1.00
    10. — Marc Andreessen (2023)0.95
    11. VorkOS0.99
    12. World'sFa0.99
    13. https://a16z.com/the-techno-optimist-manifesto/1.00
    14. Engineer1.00
    15. Id'sFai0.98
    16. Braintru0.96
    17. AlEngineer—0.87
    18. Engineering the future of Al1.00
    19. aW0.66
    20. Id'sFa0.86
    21. Engineer1.00

  • 5:03 #17 done26 line(s)

    shot 17·sharpness 4585.2

    1. Technological evolution → Growth0.99
    2. recursive1.00
    3. AlEngineer0.95
    4. World'sFair1.00
    5. "We believe that there is no material problem – whether created by nature or by0.99
    6. technology - thatcannot be solved with more technology.0.99
    7. We had a problem of starvation, so we invented the Green Revolution.0.99
    8. We had a problem of darkness,so we invented electric lighting.1.00
    9. We had a problem of cold, so we invented indoor heating.0.99
    10. We had a problem of heat, so we invented air conditioning.1.00
    11. We had a problem of isolation, so we invented the Internet.1.00
    12. We had a problem of pandemics, so we invented vaccines.0.99
    13. We have a problem of poverty, so we invent technology to create abundance."0.99
    14. AlEngi0.96
    15. —Marc Andreessen (2023)0.99
    16. Micros1.00
    17. World1.00
    18. https://a16z.com/the-techno-optimist-manifesto/1.00
    19. AlEngineer0.98
    20. Vorld'0.99
    21. Brow1.00
    22. AlEngi0.95
    23. Engineering the future of Al0.99
    24. Gra1.00
    25. World1.00
    26. AlEngine0.98

  • 5:25 #18 done71 line(s)

    shot 18·sharpness 2803.3

    1. recursive1.00
    2. AlEngineer0.98
    3. Change in one lifetime.0.99
    4. World's Fair0.98
    5. 6,0001.00
    6. Genome Project0.98
    7. 5,0001.00
    8. PCs0.90
    9. Po os0.71
    10. Man on Moon0.99
    11. 4,0001.00
    12. High Speed Computers0.99
    13. Nuclear Energy1.00
    14. 3,0001.00
    15. Peak of Greece1.00
    16. Peak of Rome1.00
    17. Discovery of0.99
    18. New World0.95
    19. Discovery of DNA1.00
    20. Invention of Airplane1.00
    21. War on Malaria0.99
    22. Penicillin1.00
    23. Black1.00
    24. 2,0001.00
    25. 1,0000.98
    26. Beginning of 1st0.99
    27. Agricultural1.00
    28. Revolution1.00
    29. Beginning of1.00
    30. Pottery0.94
    31. Invention1.00
    32. of Plow0.99
    33. 1st Irrigation1.00
    34. Works1.00
    35. 1st Cities1.00
    36. Metallurgy1.00
    37. Beginning of1.00
    38. Beginning of1.00
    39. Writing0.98
    40. Beginning of1.00
    41. Mathematics1.00
    42. Plague1.00
    43. Beginning of 2nd Agricultural Revolution1.00
    44. Beginning of 2nd Industrial Revolution0.99
    45. Invention of Watt Engine1.00
    46. Beginning of Railroads1.00
    47. Invention of Telephone Electrification1.00
    48. Germ Theory1.00
    49. AlEngineer1.00
    50. Microsoft1.00
    51. Vorld'sF0.99
    52. -9,0001.00
    53. -6,0001.00
    54. -5,0001.00
    55. -4,0001.00
    56. -3,0001.00
    57. -2,0001.00
    58. -1,0001.00
    59. 00.98
    60. 1,0000.97
    61. 2,0001.00
    62. AlEngineer0.99
    63. Fogel, R. W. (1999). Catching Up with the Economy. American Economic Review, 89(1), 1-21. https://doi.org/10.1257/aer.89.1.10.99
    64. rld's Fai0.97
    65. Browser1.00
    66. -AlEngineer -0.93
    67. Engineering the future of Al0.99
    68. adiun0.96
    69. orld'sF0.94
    70. AlEngineer1.00
    71. doil0.84

  • 6:05 #19 done37 line(s)

    shot 19·sharpness 2921.9

    1. recursive1.00
    2. AlEngineer0.99
    3. Change in one lifetime.0.99
    4. World's Fair0.96
    5. 6,0001.00
    6. Genome Project1.00
    7. 5,0001.00
    8. PRESENTEDBY1.00
    9. PCs0.99
    10. Popuin ons)0.67
    11. Man on Moon0.99
    12. Microsoft1.00
    13. 4,0001.00
    14. High Speed Computers0.98
    15. Nuclear Energy1.00
    16. 3,0001.00
    17. Peak of Greece0.97
    18. Peak of Rome1.00
    19. Discovery of0.99
    20. New World1.00
    21. Discovery of DNA1.00
    22. Invention of Airplane1.00
    23. War on Malaria1.00
    24. Penicillin0.98
    25. Black1.00
    26. 2,0001.00
    27. 1,0000.96
    28. Beginning of 1st0.98
    29. Agricultural1.00
    30. Revolution1.00
    31. Beginning of1.00
    32. Pottery0.93
    33. Invention1.00
    34. of Plow0.97
    35. 1st Irrigation1.00
    36. Works1.00
    37. 1st Cities1.00

  • 6:35 #20 skipped

    shot 20·duplicate of #19

  • 6:50 #21 done

    shot 21·sharpness 3750.4

  • 7:33 #22 done

    shot 22·sharpness 3759.8

  • 7:39 #23 done

    shot 23·sharpness 3188.9

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

149 cues· 3,143 words· 17,444 chars

  1. 0:12 All right.
  2. 0:13 Hello, everyone.
  3. 0:14 Really excited to be here.
  4. 0:15 It's a big room.
  5. 0:18 Very cool conference so far.
  6. 0:20 I want to talk to you today about something that's been on my mind for many, many years.
  7. 0:25 This is actually the first time I talk about it, sort of my version of going to Mars.
  8. 0:29 And that is the Eureka machine, a machine that will eventually invent pretty much all future inventions for humanity.
  9. 0:39 And the way we're gonna get there is by taking a step back and thinking about what else has given us a lot of really incredible inventions, namely evolution, and how that leads us to automating research and pushing the scientific frontier forward.
  10. 0:55 And this is a joint work with a lot of amazing folks at Recursive, u.com, and even some folks at AIX Ventures.
  11. 1:04 And some of these slides are actually inspired by and taken partially from one of my co-founders at Recursive, Tim Rockteschel.
  12. 1:12 So, why do I talk about evolution and why is it so important?
  13. 1:17 I think, basically, evolution is this open-ended process that has gotten us to a lot of different things that we really like.
  14. 1:26 It started in biology, it's moving to science, technology, and eventually AI, and I think it can inspire us in a lot of different ways to build better AI systems as well.
  15. 1:37 In fact, whenever we take out, and there's this famous saying, whenever I fire a linguist, my accuracy goes up.
  16. 1:44 I think that's true for machine translation back in the day.
  17. 1:48 And it may be true that we should fire all the AI engineers that are here and have them mostly manage an actual AI engineer that is AI and works on AI.
  18. 2:02 And so that may be one of the conclusions of this talk.
  19. 2:06 And I think most of us are gonna be excited about it, because it means that we'll all become managers of such an AI, rather than having to do the nitty gritty ourselves.
  20. 2:16 All right, so let's start with evolution, right?
  21. 2:18 The really, really big picture, three and a half billion years or so.
  22. 2:23 This is kind of the incredible process that has led from simple bacteria and plants and fish and amphibians and so on, to after many billions of years,
  23. 2:35 us right that's that's a good starting point that gives us some indication that evolutionary processes can do pretty amazing things right but now let's zoom in and go maybe down to a few million years there we can also see how in the very first primitive ways technological evolution
  24. 2:55 has basically increased the world's sort of product in terms of monetary value.
  25. 3:01 It's a little bit harder to estimate in the beginning, but we can see these sort of sequences of exponentials, and most exponentials eventually become S-curves, they flatten out, but humanity has done
  26. 3:14 pretty well by basically developing many of these very basic technologies, hunting, farming, but then also thinking about science, the scientific method in the early days of the Enlightenment and, of course, the Industrial Revolution.
  27. 3:27 So now we can zoom even further.
  28. 3:30 And no worries, we're eventually going to get to NanoChat and actual auto research and what we're doing.
  29. 3:34 It's a very, very quick zoom.
  30. 3:37 And now we can zoom down to the last few thousands of years, and what we're seeing there is that with more technology, we were able to sustain more people.
  31. 3:47 So when we're working on pushing that frontier forward, we're very certain that that will lead to more human flourishing.
  32. 3:54 And especially in the last few hundred years, we're seeing this incredible explosion in the population of people because of technology.
  33. 4:04 and the evolution that it brings.
  34. 4:07 And in many cases, that evolutionary process is run by us, so it's sort of conscious, but there are sort of interesting inspirations that we can take from that as we're thinking about the evolution of AI in the next cycles.
  35. 4:20 In fact, and I might not agree with everything with Marc Andreessen, but he is very smart and we agree on a lot of things.
  36. 4:27 And so I think he wrote this really great techno-optimist manifesto in which he, I think, correctly points out that the only perpetual source of growth for the entire economy, a lot of people worry about AI taking jobs and things like that, but the truth is it will very, very likely increase the economy massively and that will benefit a lot of us.
  37. 4:48 And so the perpetual source of growth is technology.
  38. 4:51 In fact, we can go even further and say that there's no material problem, and again, it's not sort of psychological problems and things like that, but no material problems that cannot be solved with even more technology.
  39. 5:04 For the problem of starvation, we invented a green revolution, darkness, light, cold, indoor heating, heat, air conditioning, and the list goes on.
  40. 5:14 So I think we can kind of realize that this evolutionary process has been going on for a very long time and continues to make a huge amount of progress.
  41. 5:23 In fact, the progress is so fast that there can, within one lifetime, be a major, major shift.
  42. 5:32 If you were born in 1900, then three years, when you're three years old, the first human ever was able to, thanks to the Wright brothers, kind of have sustained motored flight.
  43. 5:44 And then about 60-ish years later, in 1969, humans flew all the way to the moon.
  44. 5:51 So that within one lifetime, humanity went from no one can fly for a very long time, other than sort of gliding down a hill or something, no one can really fly, to we all fly to the moon.
  45. 6:04 And so for us, I think,
  46. 6:05 What that means is we are probably, and I sometimes say this, we're like too late to explore Earth, we're too early to explore the stars, but we're right on time to build an AI that could actually do what flying did for some in one lifetime due to intelligence.
  47. 6:24 We can build and move from AI being worse at everything that we do to possibly being better at any specific task that we do.
  48. 6:33 And that will probably be our 60-year timeframe, and because everything moves faster, it might only be 30 years or so.
  49. 6:40 So then there's an interesting connection between technology and science and theory, right?
  50. 6:45 Like sometimes the application comes first and then we develop the theory later and then improve the technology.

Chapters

  1. 0:00 Automating research for humanity
  2. 1:41 Why this matters now
  3. 2:19 Compressing the timeline of progress
  4. 4:28 Technoptimism and material limits
  5. 6:22 Popper and open ended evolution
  6. 9:07 The Eureka machine
  7. 10:39 Rethinking search, browsers, and GPUs
  8. 13:41 Recursive self improvement
  9. 14:46 Proof point: improving a model
  10. 16:39 Proof point: architecture search
  11. 17:16 Proof point: CUDA kernels
  12. 19:11 How far we still are, and an invitation

Open at this second