vidtheque.

read-only demo

Videos FlzpEGHNVKQ

Building a Chess Coach — Anant Dole and Asbjorn Steinskog, Take Take Take

index_state ready data_status ok

AI Engineer· published 2026-05-13· 0:18:22· en-US· indexed 2026-08-10 22:19

Open on YouTube

Scene timeline

  1. Shot 0, 0:00 to 0:05, 1 of 1 keyframes kept
  2. Shot 1, 0:05 to 0:09, 1 of 1 keyframes kept
  3. Shot 2, 0:09 to 0:14, 1 of 1 keyframes kept
  4. Shot 3, 0:14 to 0:30, 1 of 1 keyframes kept
  5. Shot 4, 0:30 to 0:49, 1 of 1 keyframes kept
  6. Shot 5, 0:49 to 1:17, 1 of 1 keyframes kept
  7. Shot 6, 1:17 to 1:29, 1 of 1 keyframes kept
  8. Shot 7, 1:29 to 1:59, 1 of 1 keyframes kept
  9. Shot 8, 1:59 to 2:29, 0 of 1 keyframes kept
  10. Shot 9, 2:29 to 2:59, 1 of 1 keyframes kept
  11. Shot 10, 2:59 to 3:27, 1 of 1 keyframes kept
  12. Shot 11, 3:27 to 3:54, 0 of 1 keyframes kept
  13. Shot 12, 3:54 to 4:43, 1 of 1 keyframes kept
  14. Shot 13, 4:43 to 5:02, 0 of 1 keyframes kept
  15. Shot 14, 5:02 to 5:06, 1 of 1 keyframes kept
  16. Shot 15, 5:06 to 5:08, 1 of 1 keyframes kept
  17. Shot 16, 5:08 to 5:18, 1 of 1 keyframes kept
  18. Shot 17, 5:18 to 5:21, 1 of 1 keyframes kept
  19. Shot 18, 5:21 to 5:22, 1 of 1 keyframes kept
  20. Shot 19, 5:22 to 5:25, 0 of 1 keyframes kept
  21. Shot 20, 5:25 to 5:26, 0 of 1 keyframes kept
  22. Shot 21, 5:26 to 5:55, 1 of 1 keyframes kept
  23. Shot 22, 5:55 to 6:24, 0 of 1 keyframes kept
  24. Shot 23, 6:24 to 6:53, 0 of 1 keyframes kept
  25. Shot 24, 6:53 to 6:55, 1 of 1 keyframes kept
  26. Shot 25, 6:55 to 7:29, 1 of 1 keyframes kept
  27. Shot 26, 7:29 to 8:04, 1 of 1 keyframes kept
  28. Shot 27, 8:04 to 8:30, 1 of 1 keyframes kept
  29. Shot 28, 8:30 to 8:55, 1 of 1 keyframes kept
  30. Shot 29, 8:55 to 9:12, 1 of 1 keyframes kept
  31. Shot 30, 9:12 to 9:57, 1 of 1 keyframes kept
  32. Shot 31, 9:57 to 10:26, 1 of 1 keyframes kept
  33. Shot 32, 10:26 to 10:55, 1 of 1 keyframes kept
  34. Shot 33, 10:55 to 11:24, 0 of 1 keyframes kept
  35. Shot 34, 11:24 to 11:44, 1 of 1 keyframes kept
  36. Shot 35, 11:44 to 12:32, 1 of 1 keyframes kept
  37. Shot 36, 12:32 to 12:37, 1 of 1 keyframes kept
  38. Shot 37, 12:37 to 12:38, 1 of 1 keyframes kept
  39. Shot 38, 12:38 to 13:19, 1 of 1 keyframes kept
  40. Shot 39, 13:19 to 13:45, 1 of 1 keyframes kept
  41. Shot 40, 13:45 to 14:12, 0 of 1 keyframes kept
  42. Shot 41, 14:12 to 14:42, 1 of 1 keyframes kept
  43. Shot 42, 14:42 to 15:12, 0 of 1 keyframes kept
  44. Shot 43, 15:12 to 15:42, 1 of 1 keyframes kept
  45. Shot 44, 15:42 to 15:53, 1 of 1 keyframes kept
  46. Shot 45, 15:53 to 16:14, 1 of 1 keyframes kept
  47. Shot 46, 16:14 to 16:37, 0 of 1 keyframes kept
  48. Shot 47, 16:37 to 17:06, 1 of 1 keyframes kept
  49. Shot 48, 17:06 to 17:34, 1 of 1 keyframes kept
  50. Shot 49, 17:34 to 17:36, 1 of 1 keyframes kept
  51. Shot 50, 17:36 to 17:38, 1 of 1 keyframes kept
  52. Shot 51, 17:38 to 17:40, 1 of 1 keyframes kept
  53. Shot 52, 17:40 to 17:51, 1 of 1 keyframes kept
  54. Shot 53, 17:51 to 18:07, 1 of 1 keyframes kept
  55. Shot 54, 18:07 to 18:21, 1 of 1 keyframes kept

55 shot(s).

keyframes kept every frame deduplicated

What was stored

cues
192
whisperx 192
chunks
31
from 192 cues
keyframes
44
kept of 55 captured
frames with text
44
1,402 lines read
chapters
0
from the source metadata
keyframe bytes
6.8 MB
word timings on 192 cues

Provenance

Each pipeline stage, its state and the model that produced it
stage state model started took
fetch done — 2026-08-10 22:15 1m 16s
stt done — 2026-08-10 22:16 26s
chunk done — 2026-08-10 22:17 0s
text_embed done — 2026-08-10 22:17 11s
keyframe done — 2026-08-10 22:17 1m 36s
ocr done — 2026-08-10 22:19 26s
frame_embed done — 2026-08-10 22:19 8s

Frames, and what the machine read

  • 0:03 #0 done2 line(s)

    shot 0·sharpness 658.4

    1. AlEngineer0.98
    2. EUROPE1.00

  • 0:08 #1 done2 line(s)

    shot 1·sharpness 829.1

    1. PRESENTINGSPONSOR1.00
    2. Google DeepMind1.00

  • 0:13 #2 done3 line(s)

    shot 2·sharpness 907.7

    1. PLATINUM SPONSORS0.98
    2. # Braintrust0.96
    3. WorkOS OpenAI0.95

  • 0:19 #3 done8 line(s)

    shot 3·sharpness 395.9

    1. Buil0.99
    2. a Chess C1.00
    3. AlEngineer1.00
    4. K0.65
    5. Anant Dole & Asbjørn Steinskog1.00
    6. Take Take Take1.00
    7. AlEngineer0.99
    8. EUROPE1.00

  • 0:41 #4 done19 line(s)

    shot 4·sharpness 1610.0

    1. Building a Chess Coach1.00
    2. AlEngineer0.98
    3. 堂0.61
    4. EUROPE1.00
    5. *★*0.59
    6. AIE1.00
    7. M0.60
    8. ★1.00
    9. ★1.00
    10. ★1.00
    11. Anant Dole & Asbjørn Steinskog1.00
    12. は0.76
    13. 80.51
    14. Take Take Take1.00
    15. AlEngineer0.97
    16. EUROPE1.00
    17. AlEngineer0.99
    18. EUROPE0.99
    19. 20250.75

  • 0:55 #5 done31 line(s)

    shot 5·sharpness 3096.8

    1. Agenda1.00
    2. 011.00
    3. What is Take Take Take?1.00
    4. Quick overview of game review • Anant0.98
    5. ★1.00
    6. 021.00
    7. A quick history of chess & Al0.99
    8. AIE1.00
    9. ★1.00
    10. Deep Blue to AlphaZero to LLMs• Asbjørn0.99
    11. ★1.00
    12. ★1.00
    13. ★1.00
    14. ★1.00
    15. 031.00
    16. Why LLMs are bad at chess0.99
    17. Kaggle competitions and chess benchmarks• Asbjørn1.00
    18. 041.00
    19. Understanding game review and closing the loop1.00
    20. Our end to end system and demo • Asbjørn0.98
    21. 051.00
    22. Latency vs. quality0.97
    23. Evaluating the trade off at scale • Anant0.98
    24. 061.00
    25. Learnings & Q&A0.97
    26. Key takeaways and open discussion• Anant0.97
    27. AlEngineer0.97
    28. EUROPE1.00
    29. AlEngineer0.97
    30. EUROPE1.00
    31. 20050.60

  • 1:20 #6 done15 line(s)

    shot 6·sharpness 402.2

    1. Agenda0.88
    2. 011.00
    3. What is Take Take Take?1.00
    4. 021.00
    5. A quick history of chess & Al0.98
    6. 031.00
    7. Why LLMs are bad at chess0.99
    8. 041.00
    9. Understanding game review and closing the0.99
    10. 051.00
    11. Latency vs. quality0.97
    12. 061.00
    13. Learnings & Q&A0.98
    14. AlEngineer0.95
    15. EUROPE1.00

  • 1:50 #7 done44 line(s)

    shot 7·sharpness 2629.9

    1. What is Take Take Take?0.99
    2. 早0.77
    3. Home1.00
    4. 13:590.90
    5. **0.82
    6. ★1.00
    7. OydeJ7140.88
    8. 2:531.00
    9. AIE1.00
    10. ★1.00
    11. ★1.00
    12. ★1.00
    13. ★1.00
    14. ★1.00
    15. ★1.00
    16. Ongoing Lichess games1.00
    17. vs Stockfish level 10.91
    18. 20.95
    19. 0100010.80
    20. Dean Yang0.95
    21. 三三0.53
    22. START A GAME0.95
    23. 三0.75
    24. 三0.85
    25. Billant moveThreat èxed Enginei I de0.71
    26. 1+00.99
    27. 3+21.00
    28. What a stunning tactial find! Moving the knight0.98
    29. discovers an attack on the g4 bishop, and if0.98
    30. 5+30.96
    31. 10+01.00
    32. Mats 11000.95
    33. 2:561.00
    34. Gyvind J gets greedy with i, they run into a0.92
    35. forced mate starting withf7+ Mats0.96
    36. effectively trades a knight for a bishop and a pawm0.97
    37. while tearing open the center0.97
    38. 10+50.93
    39. 15+101.00
    40. G0.84
    41. Engineering the future of Al1.00
    42. AlEngineer0.99
    43. EUROPE1.00
    44. 20050.72

  • 2:25 #8 skipped

    shot 8·duplicate of #7

  • 2:55 #9 done57 line(s)

    shot 9·sharpness 2891.5

    1. What is Take Take Take?1.00
    2. 早0.77
    3. Home0.99
    4. Q0.99
    5. 13:590.91
    6. 21:280.98
    7. You1.00
    8. Progress0.97
    9. Posts1.00
    10. Games0.95
    11. ★1.00
    12. ★1.00
    13. OydeJ 140.60
    14. 2:531.00
    15. ★1.00
    16. AIE1.00
    17. ★1.00
    18. ★1.00
    19. ★1.00
    20. Blitz rating1.00
    21. ★1.00
    22. ★1.00
    23. ★1.00
    24. Ongoing Lichess games1.00
    25. 10980.96
    26. vs Stockfish level 10.95
    27. 20.96
    28. 0100010.89
    29. Dean Yang0.96
    30. Accuracy1.00
    31. 20.85
    32. START A GAME0.98
    33. 三0.73
    34. Thoeat:xg4trgine: 10.63
    35. 1+01.00
    36. 3+21.00
    37. What a stunning tactical find! Moving the knight0.98
    38. Moves1.00
    39. 5+30.96
    40. 10+01.00
    41. Mats 11000.97
    42. 111.00
    43. 2:561.00
    44. Gyvind J gets greedy with xdl , they run into a0.92
    45. forced mate starting withaf7-0.96
    46. effectively trades a knight for a bishop and a pawm0.97
    47. Moves played0.99
    48. 1 5271.00
    49. 10+50.99
    50. 15+101.00
    51. Tima hraskdn0.81
    52. of0.63
    53. è0.54
    54. AlEngineer0.96
    55. EUROPE1.00
    56. AlEngineer0.99
    57. EUROPE1.00

  • 3:21 #10 done23 line(s)

    shot 10·sharpness 3377.2

    1. A quick history of Chess and AI1.00
    2. 19491.00
    3. 19971.00
    4. 20171.00
    5. 2022+1.00
    6. AIE1.00
    7. "Programming a1.00
    8. Shannon's1.00
    9. Deep Blue vs Kasparov0.98
    10. AlphaZero1.00
    11. LLMs1.00
    12. ★1.00
    13. ★1.00
    14. ★1.00
    15. Computer to0.99
    16. play chess"1.00
    17. AlphaZero1.00
    18. ChatGPT1.00
    19. AlEngineer0.97
    20. EUROPE1.00
    21. AlEngineer0.99
    22. EUROPE1.00
    23. 20250.73

  • 3:32 #11 skipped

    shot 11·duplicate of #10

  • 4:14 #12 done14 line(s)

    shot 12·sharpness 429.9

    1. A quicy0.92
    2. tory of Chess a0.99
    3. 19491.00
    4. 19971.00
    5. Shannon's1.00
    6. Deep Blue vs Kasparov1.00
    7. Al0.77
    8. K20.54
    9. "Programming a1.00
    10. Computer to0.98
    11. play chess'0.96
    12. Alp1.00
    13. AlEngineer0.99
    14. EUROPE1.00

  • 4:58 #13 skipped

    shot 13·duplicate of #10

  • 5:04 #14 done8 line(s)

    shot 14·sharpness 1983.2

    1. LLMs cannot reliably play chess1.00
    2. AIE1.00
    3. ★1.00
    4. AlEngineer0.97
    5. EUROPE1.00
    6. AlEngineer1.00
    7. EUROPE1.00
    8. 20260.83

  • 5:07 #15 done82 line(s)

    shot 15·sharpness 2278.7

    1. Chrome1.00
    2. History1.00
    3. Bookmarks1.00
    4. Profiles1.00
    5. Tab1.00
    6. Window1.00
    7. Help1.00
    8. G0.89
    9. 60.98
    10. o0.53
    11. 80.53
    12. 30.74
    13. 80.66
    14. Thu Apr 9 12:401.00
    15. [wP] AI Engineer|Bulding0.86
    16. Q Dese0.80
    17. APRIL0.90
    18. docs.google.com/presentation/d/15jhYwC8goYvWpNvtofl_JtxuBOSFiCT2Yz9_xEjkIL8/edit?slide=id.g3d4f56eae3a_0_371#slidesid.g3d4f56eae3a_0_3710.98
    19. ☆0.97
    20. G0.56
    21. 5]0.70
    22. 国0.78
    23. Work0.99
    24. # coac0.91
    25. [WIP] AI Engineer | Building a Chess Coach0.96
    26. Fil Endre Visning Sett inn Format0.98
    27. Lysbilde0.95
    28. Organiser1.00
    29. ☆@0.86
    30. Verktey0.99
    31. Utvidelser1.00
    32. Hjelp1.00
    33. 回0.57
    34. 00.76
    35. Lysbildefremvisning1.00
    36. Del0.90
    37. Messages1.00
    38. Q. Menyer0.92
    39. Tilpass1.00
    40. Animer1.00
    41. Whi1.00
    42. Sto0.84
    43. LLMs cannot reliably play chess0.99
    44. Che1.00
    45. **0.89
    46. ★1.00
    47. Anal1.00
    48. AIE1.00
    49. ★1.00
    50. Report1.00
    51. ★1.00
    52. ★1.00
    53. ★1.00
    54. ★1.00
    55. ★1.00
    56. Move:1.00
    57. Game:1.00
    58. Comm1.00
    59. Blacl0.90
    60. G0.73
    61. 690.53
    62. 口0.95
    63. from1.00
    64. com1.00
    65. e0.56
    66. O0.55
    67. Che1.00
    68. 60.65
    69. SS0.69
    70. Anal1.00
    71. B1.00
    72. Message0.99
    73. +0.93
    74. Aa1.00
    75. Gjer dette lysbildet penere ×0.95
    76. 10.14.32.030.96
    77. all0.66
    78. A0.94
    79. AlEngineer0.97
    80. EUROPE1.00
    81. AlEngineer0.99
    82. EUROPE0.99

  • 5:09 #16 done93 line(s)

    shot 16·sharpness 2368.5

    1. Edit0.93
    2. View1.00
    3. History1.00
    4. Bookmarks1.00
    5. Profiles1.00
    6. Tab1.00
    7. Window1.00
    8. Help1.00
    9. 60.98
    10. O0.62
    11. q0.52
    12. Thu Apr 9 12:401.00
    13. [WiP] AI Engineer|Build0.84
    14. APRIL0.88
    15. Q Deso0.83
    16. docs.google.com/presentation/d/15jhYwC8goYvWpNvtofl_JtxuBOSFiCT2Yz9_xEjkIL8/edit?slide=id.g3d4f56eae3a_0_371#slidesid.g3d4f56eae3a_0_3710.97
    17. ☆0.64
    18. 50.81
    19. o0.68
    20. R0.94
    21. 可0.60
    22. Work0.99
    23. # coac0.91
    24. [WIP] AI Engineer | Building a Chess Coach0.97
    25. Fil0.99
    26. Endre1.00
    27. Visning1.00
    28. Sett inn0.99
    29. Format1.00
    30. Lysbilde0.98
    31. Organiser1.00
    32. ☆0.98
    33. @0.92
    34. Verktey0.97
    35. Utvidelser1.00
    36. Hjelp1.00
    37. 回0.63
    38. 00.74
    39. Lysbidefremvisning0.98
    40. Del1.00
    41. Messages1.00
    42. Menyer1.00
    43. Tilpass1.00
    44. Animer1.00
    45. Whi1.00
    46. Sta0.84
    47. 30.84
    48. LLMs cannot reliably play chess1.00
    49. Che1.00
    50. Anal1.00
    51. ★0.98
    52. ★1.00
    53. AIE1.00
    54. ★1.00
    55. ★1.00
    56. Report1.00
    57. Ne1.00
    58. E0.69
    59. Magnus Carlsen Reacts To ChatGPT's Chess Game1.00
    60. 中0.73
    61. ★1.00
    62. ★1.00
    63. ★1.00
    64. ★1.00
    65. ★1.00
    66. Move:1.00
    67. Game:1.00
    68. Comm1.00
    69. Blacl0.93
    70. 690.52
    71. 口0.95
    72. from1.00
    73. com1.00
    74. e0.72
    75. Che1.00
    76. Sta0.94
    77. 40.74
    78. D0.60
    79. Anal1.00
    80. B1.00
    81. TAKE TAKE TAKE0.97
    82. Message1.00
    83. +0.91
    84. Aa1.00
    85. Gjer dette lysbildet penere ×0.97
    86. 器0.53
    87. 10...14.32.030.93
    88. benshot0.98
    89. A0.97
    90. AlEngineer0.97
    91. EUROPE1.00
    92. AlEngineer0.99
    93. EUROPE0.98

  • 5:21 #17 done39 line(s)

    shot 17·sharpness 1748.9

    1. www.youtube.com - To exit full screen, press0.99
    2. o30.90
    3. Grok 41.00
    4. advance will come one tempo slower.1.00
    5. 10.Bd3 à develops, but is rather slow; it doesnät0.98
    6. ★1.00
    7. ★1.00
    8. create an immediate threat.1.00
    9. *★★0.59
    10. ★★0.92
    11. 10.Kbl á also useful, but purely prophylactic.0.97
    12. AIE1.00
    13. 80.99
    14. ★1.00
    15. 10.g4 à the main-line move in this position. It0.99
    16. ★1.00
    17. ★1.00
    18. ★1.00
    19. supports f5, grabs space, can push g5, and starts a1.00
    20. direct assault while Blackäs king is still in the0.99
    21. middle.1.00
    22. 80.98
    23. 80.99
    24. 10.g4 is the theoretical main move and fits0.97
    25. perfectly with Whiteás strategic aims. It is0.98
    26. completely sound, gains space, and prepares f51.00
    27. next move.1.00
    28. Therefore the strongest move is:1.00
    29. [Grok 4 (black) thoughts]1.00
    30. Final Answer: g41.00
    31. Final Answer: b50.99
    32. kaggle0.99
    33. TAKETAKETAKE1.00
    34. THE BEST APP TO FOLLOW CHESS1.00
    35. AlEngineer0.97
    36. EUROPE1.00
    37. AlEngineer0.99
    38. EUROPE0.99
    39. 20260.83

  • 5:22 #18 done17 line(s)

    shot 18·sharpness 1740.8

    1. Magnus Carlsen Reacts To ChatGPT's Chess Game0.99
    2. www.youtube.com- To exit full screen, press0.97
    3. ***0.75
    4. AIE1.00
    5. ★0.99
    6. ★1.00
    7. ★1.00
    8. ★*0.75
    9. TYLE1.00
    10. ESS1.00
    11. 3320.72
    12. Se påYouTube0.93
    13. AlEngineer0.96
    14. EUROPE1.00
    15. AlEngineer0.99
    16. EUROPE1.00
    17. 20260.86

  • 5:25 #19 skipped

    shot 19·duplicate of #15

  • 5:26 #20 skipped

    shot 20·duplicate of #14

  • 5:33 #21 done12 line(s)

    shot 21·sharpness 2831.4

    1. LLMs cannot reliably play chess0.99
    2. Magnus Carlsen Reacts To ChatGPT's Chess Game1.00
    3. AIE1.00
    4. ★1.00
    5. ★1.00
    6. TYLE1.00
    7. ESS1.00
    8. Se pYouTube0.97
    9. Engineering the future of Al1.00
    10. AlEngineer1.00
    11. EUROPE1.00
    12. 20260.80

  • 6:18 #22 skipped

    shot 22·duplicate of #21

  • 6:44 #23 skipped

    shot 23·duplicate of #21

Transcript

192 cues· 3,041 words· 16,350 chars

  1. 0:14 Afternoon, everyone.
  2. 0:16 So our next talk will be something a little bit different.
  3. 0:20 We're going to dive into the world of chess.
  4. 0:23 Quick show of hands.
  5. 0:24 Who has heard of Magnus Carlsen?
  6. 0:27 Okay, fantastic.
  7. 0:29 No introduction needed, but widely considered the best chess player in the world.
  8. 0:33 He also founded a company called TakeTakeTake.
  9. 0:37 This is where myself, Anant, and my colleague, Aspern, currently work at.
  10. 0:42 And we're gonna talk to you today about how we built our AI chess coach that now you can use and is in production.
  11. 0:50 So first up, quick agenda.
  12. 0:52 We'll quickly discuss a bit more about TakeTakeTake, what it is we actually built, what it is we actually launched.
  13. 0:58 Aspet will then go into a quick history of chess and AI, a lot of links there.
  14. 1:02 We'll briefly touch on why LLMs are actually bad at chess and how we managed to solve this problem.
  15. 1:08 We're then gonna deep dive into actually understanding our game review and sort of closing the loop with our autonomous agent, and you'll get a demo.
  16. 1:15 And then finally, some latency versus quality trade-offs, as this is a consumer-focused AI application.
  17. 1:22 And then lastly, some learnings.
  18. 1:25 So first up, what is Tech Tech Take?
  19. 1:27 In its simplest form today, it's currently an iOS and Android application.
  20. 1:31 You can go on and play your friends, and you can post about your games.
  21. 1:36 What's relevant for our particular talk is that after you play a game, you get presented with our game review.
  22. 1:42 And this is powered by our AI pipeline.
  23. 1:46 So for example, just showing you how it works, in this particular position,
  24. 1:50 It's leading to a checkmate.
  25. 1:53 The last move that white has played has moved the knight from this yellow square over here on f3, captured the pawn on e5.
  26. 2:00 It is a brilliant move, so automatically gets the brilliant sort of notation.
  27. 2:04 And the commentary below is actually generated by our system.
  28. 2:08 And we're using an LLM, and the pipeline we'll get into in a second.
  29. 2:12 But what's quite interesting about it is we're able to give you the nuance of why it is a tactic, what detectors from a positional and tactical sense have fired, what are the threats you're trying to do, and actually explain sort of the why behind the move.
  30. 2:27 So that's the system we're going to be talking about today.
  31. 2:31 Finally, on the last of the step of our application, we've started revealing insights about your play.
  32. 2:38 And this could be things like how accurate you played in a particular game phase, maybe your current rating, or your current depth in a particular opening.
  33. 2:46 And these insights form the next layer of analysis that we present to the coach, who then gives them back to you as opportunities for learning and improving.
  34. 2:54 We hope by using this, you'll be able to improve and become better at the game.
  35. 2:59 All right.
  36. 3:00 So first, a brief history of chess and AI since they've been intertwined for so long.
  37. 3:05 Just to give you a little bit of a back story.
  38. 3:07 In 1949, Claude Shannon, the OG Claude, wrote the paper Programming a Computer to Play Chess.
  39. 3:14 And here he envisioned that, or he proposed that there are two types of chess engines, type A and type B.
  40. 3:22 Type A were these brute force engines that search through all possible moves and figure out the best move, while Type B were those who we know from 2017 and onward that can selectively pick out the best moves.
  41. 3:40 Back then, he assumed that we would need Type B computers to play chess because computers were so weak back then, you couldn't search through the whole tree of moves.
  42. 3:51 But computers quickly became better, and people just started scaling these type A computers.
  43. 3:58 They got better and better until they, in 1997, Deep Blue versus Kasparov, the first time a chess engine beat the best chess player at the time.
  44. 4:09 So people didn't really bother about these type B computers for a while, these intuitive engines, until
  45. 4:18 DeepMind, shout out to DeepMind, released first AlphaGo, because Go is a much more complex game than chess.
  46. 4:26 So you can't solve this with these type A computers.
  47. 4:29 You would need this intuitive approach, neural network approach, to actually selectively figure out which lines to calculate.
  48. 4:35 But after that, they released AlphaZero, who could play not only Go, but also chess and shogi.
  49. 4:44 some some years later, LMS came and people started playing chess against the LMS and quickly turned out that they can't really play chess.
  50. 4:52 Sometimes they they make some right moves and they can't to an extent, play play a nice opening, but they quickly start to hallucinate.

Open at this second