vidtheque.

read-only demo

Videos XKup1pj-34M

The New Application Layer - Malte Ubl, CTO Vercel

index_state ready data_status ok

AI Engineer· published 2026-04-20· 0:18:51· en-US· indexed 2026-08-10 19:52

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 0:20, 1 of 1 keyframes kept
  5. Shot 4, 0:20 to 0:42, 1 of 1 keyframes kept
  6. Shot 5, 0:42 to 0:44, 1 of 1 keyframes kept
  7. Shot 6, 0:44 to 0:49, 1 of 1 keyframes kept
  8. Shot 7, 0:49 to 0:51, 1 of 1 keyframes kept
  9. Shot 8, 0:51 to 0:53, 1 of 1 keyframes kept
  10. Shot 9, 0:53 to 0:56, 1 of 1 keyframes kept
  11. Shot 10, 0:56 to 1:02, 1 of 1 keyframes kept
  12. Shot 11, 1:02 to 1:36, 1 of 1 keyframes kept
  13. Shot 12, 1:36 to 1:51, 1 of 1 keyframes kept
  14. Shot 13, 1:51 to 2:22, 1 of 1 keyframes kept
  15. Shot 14, 2:22 to 3:00, 1 of 1 keyframes kept
  16. Shot 15, 3:00 to 3:24, 1 of 1 keyframes kept
  17. Shot 16, 3:24 to 3:55, 1 of 1 keyframes kept
  18. Shot 17, 3:55 to 4:07, 1 of 1 keyframes kept
  19. Shot 18, 4:07 to 4:56, 1 of 1 keyframes kept
  20. Shot 19, 4:56 to 5:21, 1 of 1 keyframes kept
  21. Shot 20, 5:21 to 6:04, 1 of 1 keyframes kept
  22. Shot 21, 6:04 to 6:46, 1 of 1 keyframes kept
  23. Shot 22, 6:46 to 6:57, 1 of 1 keyframes kept
  24. Shot 23, 6:57 to 7:23, 1 of 1 keyframes kept
  25. Shot 24, 7:23 to 7:49, 1 of 1 keyframes kept
  26. Shot 25, 7:49 to 8:15, 0 of 1 keyframes kept
  27. Shot 26, 8:15 to 8:41, 1 of 1 keyframes kept
  28. Shot 27, 8:41 to 9:07, 1 of 1 keyframes kept
  29. Shot 28, 9:07 to 9:33, 1 of 1 keyframes kept
  30. Shot 29, 9:33 to 9:59, 1 of 1 keyframes kept
  31. Shot 30, 9:59 to 10:25, 1 of 1 keyframes kept
  32. Shot 31, 10:25 to 10:51, 1 of 1 keyframes kept
  33. Shot 32, 10:51 to 11:17, 0 of 1 keyframes kept
  34. Shot 33, 11:17 to 11:42, 1 of 1 keyframes kept
  35. Shot 34, 11:42 to 12:08, 0 of 1 keyframes kept
  36. Shot 35, 12:08 to 12:34, 0 of 1 keyframes kept
  37. Shot 36, 12:34 to 12:49, 1 of 1 keyframes kept
  38. Shot 37, 12:49 to 13:07, 1 of 1 keyframes kept
  39. Shot 38, 13:07 to 13:35, 1 of 1 keyframes kept
  40. Shot 39, 13:35 to 14:04, 1 of 1 keyframes kept
  41. Shot 40, 14:04 to 14:32, 1 of 1 keyframes kept
  42. Shot 41, 14:32 to 15:00, 1 of 1 keyframes kept
  43. Shot 42, 15:00 to 15:29, 1 of 1 keyframes kept
  44. Shot 43, 15:29 to 15:57, 1 of 1 keyframes kept
  45. Shot 44, 15:57 to 16:20, 1 of 1 keyframes kept
  46. Shot 45, 16:20 to 16:29, 1 of 1 keyframes kept
  47. Shot 46, 16:29 to 17:09, 1 of 1 keyframes kept
  48. Shot 47, 17:09 to 17:14, 1 of 1 keyframes kept
  49. Shot 48, 17:14 to 17:42, 1 of 1 keyframes kept
  50. Shot 49, 17:42 to 18:11, 1 of 1 keyframes kept
  51. Shot 50, 18:11 to 18:19, 1 of 1 keyframes kept
  52. Shot 51, 18:19 to 18:32, 1 of 1 keyframes kept
  53. Shot 52, 18:32 to 18:35, 1 of 1 keyframes kept
  54. Shot 53, 18:35 to 18:36, 1 of 1 keyframes kept
  55. Shot 54, 18:36 to 18:50, 1 of 1 keyframes kept
  56. Shot 55, 18:50 to 18:51, 0 of 1 keyframes kept

56 shot(s).

keyframes kept every frame deduplicated

What was stored

cues
225
whisperx 225
chunks
33
from 225 cues
keyframes
51
kept of 56 captured
frames with text
51
1,091 lines read
chapters
0
from the source metadata
keyframe bytes
5.9 MB
word timings on 225 cues

Provenance

Each pipeline stage, its state and the model that produced it
stage state model started took
fetch done — 2026-08-10 16:41 1m 49s
stt done — 2026-08-10 16:43 20s
chunk done — 2026-08-10 16:43 0s
text_embed done — 2026-08-10 19:52 1s
keyframe done — 2026-08-10 16:43 1m 40s
ocr done — 2026-08-10 16:45 18s
frame_embed done — 2026-08-10 19:52 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

  • 0:15 #3 done1 line(s)

    shot 3·sharpness 107.0

    1. AIF0.95

  • 0:31 #4 done6 line(s)

    shot 4·sharpness 419.4

    1. AlEngineer0.99
    2. ▲Vercel1.00
    3. AlEngineer0.99
    4. EUROPE1.00
    5. EUROPE1.00
    6. AIF0.91

  • 0:44 #5 done3 line(s)

    shot 5·sharpness 708.2

    1. MALTE1.00
    2. UBL1.00
    3. ▲Vercel0.99

  • 0:46 #6 done5 line(s)

    shot 6·sharpness 490.4

    1. MALTE1.00
    2. AIE1.00
    3. UBL0.86
    4. CTO1.00
    5. AIE1.00

  • 0:49 #7 done6 line(s)

    shot 7·sharpness 320.3

    1. UBL0.95
    2. CTO1.00
    3. AlEngineer0.99
    4. EUROPE1.00
    5. PRESENTED BY1.00
    6. Google DeepMind1.00

  • 0:53 #8 done29 line(s)

    shot 8·sharpness 996.6

    1. UDL1.00
    2. OPE1.00
    3. ettenaii0.96
    4. AlEngijieerr0.93
    5. EUROPP0.99
    6. CTO1.00
    7. Brainttustt0.93
    8. AMEngijieeer0.83
    9. eepMirad0.68
    10. AEEnjineer0.88
    11. EUROPPE0.96
    12. ELBUROPE0.93
    13. :neo4j0.90
    14. AAEF0.80
    15. Mirasfft0.76
    16. Verce1.00
    17. er1.00
    18. BY0.76
    19. Trigtbew0.65
    20. AEEsgijiseeer0.82
    21. HOS0.76
    22. AAEEagjireeer0.83
    23. Mind1.00
    24. EURROIPE0.80
    25. EUROEPE0.88
    26. AAEE0.96
    27. ttailbscate0.74
    28. AE0.66
    29. Siurorr0.68

  • 0:56 #9 done43 line(s)

    shot 9·sharpness 1179.5

    1. EUROPE1.00
    2. together.ai1.00
    3. AlEngineer0.97
    4. AlEngin0.95
    5. EUROPE1.00
    6. EUROPE1.00
    7. AlEngineer0.99
    8. Braintrust1.00
    9. Google DeepMind1.00
    10. AlEngineer0.97
    11. Open1.00
    12. EUROPE1.00
    13. EUROPE1.00
    14. arize1.00
    15. AlEngineer0.99
    16. M1.00
    17. AlEngineer1.00
    18. neo4j0.98
    19. AIEngine0.95
    20. EUROPE1.00
    21. EUROPE1.00
    22. EUROPE1.00
    23. eer1.00
    24. BY0.99
    25. AlEngineer0.96
    26. Trigger.dev1.00
    27. AIE1.00
    28. WorkOS1.00
    29. AlEngineer0.96
    30. Tes0.90
    31. epMind1.00
    32. EUROPE1.00
    33. EUROPE1.00
    34. Modal1.00
    35. AlEngineer0.99
    36. ta1.00
    37. :al0.96
    38. AlEngineer1.00
    39. Sonar1.00
    40. AlEngine0.93
    41. EUROPE1.00
    42. EUROPE1.00
    43. EUROPE1.00

  • 0:59 #10 done20 line(s)

    shot 10·sharpness 1252.2

    1. enLubs0.82
    2. ngineer1.00
    3. AI ENGINEER LONDON1.00
    4. EUROPE1.00
    5. The new application layer1.00
    6. aintru0.98
    7. ngineer0.93
    8. ngineer1.00
    9. Vicrosoft0.96
    10. EUROPE1.00
    11. gger.de0.99
    12. Engineer0.95
    13. EUROPE1.00
    14. ngineer1.00
    15. ilscale1.00
    16. EUROPE1.00
    17. MALTE UBL0.98
    18. ▲VERCEL CTO0.96
    19. AlEngineer0.97
    20. EUROPE1.00

  • 1:06 #11 done23 line(s)

    shot 11·sharpness 1550.3

    1. EUROP1.00
    2. AlEngin0.96
    3. Deep1.00
    4. EUROPE1.00
    5. Micros1.00
    6. AlEngineer0.98
    7. EUROPE1.00
    8. Chat1.00
    9. SDK1.00
    10. JUST BASH0.99
    11. AlEngineel0.94
    12. Work1.00
    13. EUROPE1.00
    14. tailsca1.00
    15. AlEngineer0.92
    16. EUROPE1.00
    17. AlEngineer0.98
    18. THE NEW APPLICATION LAYER1.00
    19. EUROPE1.00
    20. PRESENTED BY1.00
    21. Google DeepMind1.00
    22. MALTE UBL / CTO0.91
    23. ▲Vercel0.98

  • 1:41 #12 done17 line(s)

    shot 12·sharpness 3235.2

    1. togetherai0.97
    2. AlEngineer0.99
    3. norkel0.98
    4. EUROPE1.00
    5. SENTRY1.00
    6. eer0.99
    7. AlEngineer0.99
    8. :rosol0.84
    9. EUROPE1.00
    10. eNCORD0.96
    11. AlEngineer0.99
    12. THE NEW APPLICATION LAYER1.00
    13. EUROPE1.00
    14. PRESENTED BY1.00
    15. Google DeepMind1.00
    16. MALTE UBL / CTO0.97
    17. ▲Vercel1.00

  • 1:57 #13 done22 line(s)

    shot 13·sharpness 3232.7

    1. EUROF0.91
    2. Welcn0.93
    3. ome to0.91
    4. AlEnginee0.97
    5. Deep0.93
    6. EUROPE1.00
    7. Microsoa0.89
    8. AlEnginee0.96
    9. EUROPE1.00
    10. AlEngineer0.97
    11. Work1.00
    12. EUROPE1.00
    13. tailsc1.00
    14. AlEnginee0.98
    15. EUROPE1.00
    16. AlEngineer0.98
    17. THE NEW APPLICATION LAYER1.00
    18. EUROPE1.00
    19. PRESENTED BY1.00
    20. Google DeepMind1.00
    21. MALTE UBL / CTO0.92
    22. ▲Vercel0.98

  • 2:48 #14 done26 line(s)

    shot 14·sharpness 1655.8

    1. AlEngineer0.99
    2. togethers0.93
    3. EUROPE1.00
    4. Ope0.98
    5. AlEngineer0.95
    6. EUROPE1.00
    7. From pages to agents0.98
    8. AlEn0.95
    9. SENTRY0.93
    10. SITES1.00
    11. APPS1.00
    12. +AGENTS0.95
    13. EUR0.85
    14. T0.95
    15. AlEngineer0.96
    16. EUROPE0.99
    17. BeNCORD0.92
    18. AlEngin0.98
    19. EUROPE1.00
    20. AlEngineer0.98
    21. THE NEW APPLICATION LAYER1.00
    22. EUROPE1.00
    23. PRESENTED BY1.00
    24. Google DeepMind1.00
    25. MALTE UBL /CTO0.95
    26. ▲Vercel0.99

  • 3:07 #15 done25 line(s)

    shot 15·sharpness 1830.0

    1. EUROPE1.00
    2. THEDISRUPTION1.00
    3. Enginee0.93
    4. WorkOS1.00
    5. EUROPE1.00
    6. eo4j0.92
    7. AlEngineer0.97
    8. EUROPE1.00
    9. Engineer1.00
    10. Braintrust0.99
    11. EUROPE1.00
    12. How we build0.95
    13. What we build0.98
    14. ailscale1.00
    15. IEngineer0.98
    16. EUROPE1.00
    17. WITH AI AGENTS0.99
    18. AI AGENTS1.00
    19. AlEngineer0.98
    20. THE NEW APPLICATION LAYER1.00
    21. EUROPE1.00
    22. PRESENTED BY1.00
    23. Google DeepMind1.00
    24. MALTE UBL/CTO0.98
    25. ▲Vercel1.00

  • 3:37 #16 done19 line(s)

    shot 16·sharpness 1422.8

    1. EURC0.93
    2. AlEngi0.83
    3. Wo1.00
    4. neo1.00
    5. AIEngi0.94
    6. EUROP0.98
    7. AIEngine0.95
    8. Brail0.97
    9. EUROPE1.00
    10. tailsca1.00
    11. AlEngi0.94
    12. EUROF1.00
    13. AlEngineer0.98
    14. THE NEW APPLICATION LAYER1.00
    15. EUROPE1.00
    16. PRESENTED BY1.00
    17. Google DeepMind1.00
    18. MALTE UBL /CTO0.94
    19. ▲Vercel1.00

  • 4:02 #17 done22 line(s)

    shot 17·sharpness 1615.3

    1. Brain0.92
    2. AlEngineer1.00
    3. EUROPE0.99
    4. o4j0.87
    5. AI0.91
    6. Agents are a new0.97
    7. Engineer1.00
    8. kind of software0.99
    9. EUROPE1.00
    10. Tessl1.00
    11. Al0.81
    12. Sonc0.91
    13. IEngineer0.95
    14. EUROPE0.99
    15. AlEngineer0.98
    16. THE NEW APPLICATION LAYER1.00
    17. EUROPE1.00
    18. PRESENTED BY1.00
    19. Google DeepMind1.00
    20. Braintrust1.00
    21. WorkOS1.00
    22. OpenAl0.94

  • 4:45 #18 done23 line(s)

    shot 18·sharpness 1709.4

    1. EUROF0.99
    2. AlEngine0.95
    3. Work1.00
    4. EUROPE1.00
    5. ne040.87
    6. AlEnginee0.98
    7. EUROPE1.00
    8. SOFTWARE THAT WOULD BE NICE TO HAVE0.98
    9. AlEngine0.97
    10. : Braintr0.92
    11. ECONOMICALLY VIABLE TRADITIONAL SOFTWARE1.00
    12. EUROPE1.00
    13. tailsc0.99
    14. AlEnginee0.99
    15. EUROPE1.00
    16. AlEngineer0.96
    17. THE NEW APPLICATION LAYER1.00
    18. EUROPE1.00
    19. PRESENTED BY1.00
    20. Google DeepMind1.00
    21. Braintrust1.00
    22. WorkOS1.00
    23. OpenAl0.94

  • 5:06 #19 done19 line(s)

    shot 19·sharpness 1457.4

    1. Engineer1.00
    2. EUROPE1.00
    3. eor0.81
    4. IEngineer0.99
    5. EUROPE1.00
    6. Engineer1.00
    7. OpenAI0.96
    8. SOFTWARE THAT WOULD BE NICE TO HAVE1.00
    9. EUROPE1.00
    10. ECONOMICALLY VIABLE TRADITIONAL SOFTWARE1.00
    11. SENTF1.00
    12. AlEngineer0.98
    13. EUROPE1.00
    14. ECONOMICALLY VIABLE AGENTS0.98
    15. Engineer1.00
    16. Modal1.00
    17. EUROPE1.00
    18. Engineering the future of Al0.98
    19. AlEngineer0.99

  • 5:59 #20 done15 line(s)

    shot 20·sharpness 1139.2

    1. er1.00
    2. eer1.00
    3. epMind1.00
    4. BUY1.00
    5. MAKE1.00
    6. eer1.00
    7. BY1.00
    8. pMind1.00
    9. AlEngineer0.98
    10. THE NEW APPLICATION LAYER1.00
    11. EUROPE1.00
    12. PRESENTED BY0.99
    13. Google DeepMind1.00
    14. MALTE UBL / CTO0.93
    15. ▲Vercel0.99

  • 6:09 #21 done59 line(s)

    shot 21·sharpness 2701.7

    1. AlEngin0.98
    2. Job postings for software engineers are rapidly rising1.00
    3. EUROPE1.00
    4. Indeed Job Postings: Software Engineers (LHS) + 21dma1.00
    5. Overall Indeed Job Postings (RHS) + 21dma0.99
    6. All1.00
    7. Wor1.00
    8. 751.00
    9. 741.00
    10. 1251.00
    11. 731.00
    12. 721.00
    13. 1201.00
    14. 711.00
    15. 701.00
    16. 691.00
    17. 1150.99
    18. Google1.00
    19. AlEngin0.98
    20. 681.00
    21. 671.00
    22. EUROPE1.00
    23. 661.00
    24. 1101.00
    25. 651.00
    26. 641.00
    27. AlEngineer1.00
    28. 631.00
    29. 1051.00
    30. EUROPE1.00
    31. Λari0.92
    32. 621.00
    33. 611.00
    34. 601.00
    35. 1001.00
    36. PRESENTED BY1.00
    37. Google DeepMind0.98
    38. JAN 241.00
    39. MAR 241.00
    40. MAY 241.00
    41. JUL 241.00
    42. SEP 241.00
    43. NOV 241.00
    44. JAN 251.00
    45. MAR 251.00
    46. MAY 251.00
    47. JUL 250.92
    48. SEP 251.00
    49. NOV 251.00
    50. JAN 260.96
    51. AlEngin0.96
    52. EUROPE1.00
    53. AlEngineer0.97
    54. THE NEW APPLICATION LAYER1.00
    55. EUROPE1.00
    56. PRESENTED BY1.00
    57. Google DeepMind1.00
    58. MALTE UBL /CTO0.96
    59. ▲Vercel0.96

  • 6:47 #22 done22 line(s)

    shot 22·sharpness 1663.9

    1. EURO1.00
    2. AlEngine0.94
    3. Brain1.00
    4. EUROPE1.00
    5. OpenA1.00
    6. AlEngin0.94
    7. Let's build the next0.96
    8. EUROPE1.00
    9. application layer1.00
    10. AIEngine0.94
    11. Micr1.00
    12. EUROPE1.00
    13. Mor0.94
    14. AIEngin0.94
    15. EUROPE1.00
    16. AlEngineer0.98
    17. THE NEW APPLICATION LAYER1.00
    18. EUROPE1.00
    19. PRESENTED BY1.00
    20. Google DeepMind1.00
    21. MALTE UBL / CTO0.92
    22. ▲Vercel1.00

  • 7:13 #23 done26 line(s)

    shot 23·sharpness 1860.0

    1. AlEngineer0.98
    2. EUROPE1.00
    3. SA0.78
    4. AlEngineer0.96
    5. INTEL1.00
    6. EUROPE1.00
    7. Archetypes1.00
    8. AlEngine0.96
    9. ioogle Deep0.99
    10. of practical0.96
    11. EUROPE1.00
    12. non-coding1.00
    13. right1.00
    14. AlEngineer0.92
    15. EUROPE0.96
    16. of agents0.96
    17. AlEngin0.94
    18. eNCOF0.92
    19. EUROPE1.00
    20. AlEngineer0.98
    21. THE NEW APPLICATION LAYER1.00
    22. EUROPE1.00
    23. PRESENTED BY1.00
    24. Google DeepMind1.00
    25. MALTE UBL/CTO0.98
    26. ▲Vercel0.97

Transcript

225 cues· 3,038 words· 16,593 chars

  1. 0:15 The first speaker draws on over 25 years of software engineering experience.
  2. 0:22 From his time at Google and now Vercel, he will explore what it means to build infrastructure and applications in a world where agents are both the builders and users of software.
  3. 0:35 Please join me in welcoming to the stage the CTO of Vercel, Malte Ubel.
  4. 0:50 Good morning, everyone.
  5. 0:52 This is awesome.
  6. 0:53 I'm so glad to be here.
  7. 0:54 Welcome to the first ever AI engineer conference in Europe.
  8. 0:58 My name is Malte, and I'm the CTO of Vercel.
  9. 1:02 Now, usually I give technical talks, but I thought, because I'm currently going first, that I need to give a proper keynote.
  10. 1:10 But I didn't want to feature what I call my vibe coding stack.
  11. 1:15 I've been hacking on the thing called chat SDK, which is a way to hook your agents to whatever, like Slack, Telegram, WhatsApp.
  12. 1:22 chat app you like, and I've been working on Just Bash, which is a Bash interpreter written in TypeScript that gives you something like a sandbox with nanosecond startup time for your agents because they love Bash.
  13. 1:35 All right.
  14. 1:39 One thing I wanted to mention is that
  15. 1:41 The reason why I'm so excited to be here is that I used to run a little conference in Berlin called JSConf EU.
  16. 1:47 And I feel like once in my life I had completely impeccable timing because it was the summer of 2019 and we decided after 10 years it was enough and we went out with a bang.
  17. 1:59 And the reason why this was such great timing was that there just wouldn't have been a conference one year later because of COVID.
  18. 2:07 But also when we decided that we would step away, we were hoping that someone else would take the reins.
  19. 2:12 And again, that did not happen because of COVID.
  20. 2:16 So it's now been more than half a decade, and I'm very excited that things are finally starting up again.
  21. 2:23 But it was also clear that it wasn't going to be like a web development conference that would really bring the tech community in Europe back together in 2026.
  22. 2:32 In many ways, I think AI engineering is the legitimate successor to web development as a really mainstream discipline of engineering that will shape the next decade of software development as software engineering itself faces an unprecedented disruption.
  23. 2:50 So you're definitely in the right place today.
  24. 2:52 And it's more important than ever to come together as a community and reflect on both our profession as software engineers and AI engineers.
  25. 3:01 And that's because we're facing a disruption of both how we build, which is with AI, and what we build, which is AI and agents.
  26. 3:11 And of course, disruption can sometimes lead to anxiety.
  27. 3:14 In fact, I really actually very often get asked, hey Malte, is there still a place for engineers in the future?
  28. 3:22 And what about that next generation of engineers?
  29. 3:26 And I couldn't be more convinced that the answer is yes.
  30. 3:30 I often give this example of envision me doing a TikTok video.
  31. 3:34 They mentioned in the intro that I have 25 years of experience, which is actually a substantial understatement.
  32. 3:39 And so I would not be good at a TikTok video.
  33. 3:42 I should not be recording TikTok videos because I didn't grow up with this.
  34. 3:46 And in a very similar way, the next generation of junior engineers are just going to be so much better at this discipline because they get molded in the AI world, just like all of you are.
  35. 3:57 But it's not only the kids that are going to be all right.
  36. 3:59 Well, I'll be fine.
  37. 4:00 And this is why.
  38. 4:03 One of our main see this is that agents are a new kind of software.
  39. 4:08 Because there was always all this stuff we wanted to automate.
  40. 4:12 But not all of it was economically viable to do with traditional software.
  41. 4:17 But it is with agents.
  42. 4:19 And what that means is there will be just so much more software in the future.
  43. 4:23 Indulge me with a Venn diagram.
  44. 4:27 Maybe the circle should be better because the circle represents all software that should exist.
  45. 4:31 Imagine all software that should exist.
  46. 4:36 The problem was that we couldn't write all of it because it was too expensive using traditional Mexit.
  47. 4:43 You can envision all these things where you have all these if statements, you have all this knowledge about the business, you have to figure it all out, you have to hard code into the application.
  48. 4:51 So much of this software you just would never write because it would be obviously too expensive.
  49. 4:58 But now with agents, that part of software becomes economically viable.
  50. 5:03 I can build it now with not that much effort.

Open at this second