vidtheque.

read-only demo

Videos ugUeZ8-b-u0

Why Rust is the Ideal Language for Vibe-Coding — Daniel Szoke, Sentry

index_state ready data_status ok

AI Engineer· published 2026-05-27· 0:16:24· en-US· indexed 2026-08-10 19:56

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 0:41, 1 of 1 keyframes kept
  5. Shot 4, 0:41 to 1:07, 1 of 1 keyframes kept
  6. Shot 5, 1:07 to 1:33, 1 of 1 keyframes kept
  7. Shot 6, 1:33 to 1:59, 1 of 1 keyframes kept
  8. Shot 7, 1:59 to 2:24, 1 of 1 keyframes kept
  9. Shot 8, 2:24 to 2:49, 0 of 1 keyframes kept
  10. Shot 9, 2:49 to 3:14, 0 of 1 keyframes kept
  11. Shot 10, 3:14 to 3:40, 1 of 1 keyframes kept
  12. Shot 11, 3:40 to 4:05, 0 of 1 keyframes kept
  13. Shot 12, 4:05 to 4:30, 1 of 1 keyframes kept
  14. Shot 13, 4:30 to 4:55, 0 of 1 keyframes kept
  15. Shot 14, 4:55 to 5:21, 0 of 1 keyframes kept
  16. Shot 15, 5:21 to 5:46, 1 of 1 keyframes kept
  17. Shot 16, 5:46 to 6:11, 0 of 1 keyframes kept
  18. Shot 17, 6:11 to 6:36, 0 of 1 keyframes kept
  19. Shot 18, 6:36 to 7:01, 1 of 1 keyframes kept
  20. Shot 19, 7:01 to 7:27, 0 of 1 keyframes kept
  21. Shot 20, 7:27 to 7:52, 1 of 1 keyframes kept
  22. Shot 21, 7:52 to 8:17, 0 of 1 keyframes kept
  23. Shot 22, 8:17 to 8:42, 1 of 1 keyframes kept
  24. Shot 23, 8:42 to 9:08, 0 of 1 keyframes kept
  25. Shot 24, 9:08 to 9:33, 0 of 1 keyframes kept
  26. Shot 25, 9:33 to 9:58, 1 of 1 keyframes kept
  27. Shot 26, 9:58 to 10:23, 0 of 1 keyframes kept
  28. Shot 27, 10:23 to 10:49, 0 of 1 keyframes kept
  29. Shot 28, 10:49 to 11:14, 0 of 1 keyframes kept
  30. Shot 29, 11:14 to 11:39, 0 of 1 keyframes kept
  31. Shot 30, 11:39 to 12:04, 1 of 1 keyframes kept
  32. Shot 31, 12:04 to 12:30, 0 of 1 keyframes kept
  33. Shot 32, 12:30 to 12:55, 0 of 1 keyframes kept
  34. Shot 33, 12:55 to 13:20, 0 of 1 keyframes kept
  35. Shot 34, 13:20 to 13:48, 1 of 1 keyframes kept
  36. Shot 35, 13:48 to 14:16, 1 of 1 keyframes kept
  37. Shot 36, 14:16 to 14:44, 1 of 1 keyframes kept
  38. Shot 37, 14:44 to 15:13, 1 of 1 keyframes kept
  39. Shot 38, 15:13 to 15:41, 1 of 1 keyframes kept
  40. Shot 39, 15:41 to 16:09, 1 of 1 keyframes kept
  41. Shot 40, 16:09 to 16:24, 1 of 1 keyframes kept

41 shot(s).

keyframes kept every frame deduplicated

What was stored

cues
124
whisperx 124
chunks
27
from 124 cues
keyframes
23
kept of 41 captured
frames with text
23
310 lines read
chapters
10
from the source metadata
keyframe bytes
5.2 MB
word timings on 124 cues

Provenance

Each pipeline stage, its state and the model that produced it
stage state model started took
fetch done — 2026-08-10 18:56 2m 19s
stt done — 2026-08-10 18:58 16s
chunk done — 2026-08-10 18:58 0s
text_embed done — 2026-08-10 19:56 0s
keyframe done — 2026-08-10 18:58 1m 29s
ocr done — 2026-08-10 19:00 11s
frame_embed done — 2026-08-10 19:56 4s

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 907.6

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

  • 0:25 #3 done11 line(s)

    shot 3·sharpness 1764.8

    1. London, April 20261.00
    2. Why Rust is the Ideal1.00
    3. *★★0.64
    4. Language for Vibe-1.00
    5. AIE1.00
    6. Coding1.00
    7. Daniel Szoke0.99
    8. Why Rust is the Ideal1.00
    9. Language for Vibe-1.00
    10. Coding1.00
    11. Google DeepMind1.00

  • 0:56 #4 done36 line(s)

    shot 4·sharpness 2333.6

    1. Private1.00
    2. chatgpt.com1.00
    3. ChatGPT1.00
    4. Log in0.97
    5. Sign up for free1.00
    6. In 2026, vibe coding = Al-first development where you describe intent and the model writes0.99
    7. Q1.00
    8. code, but you still guide, review, and fix it.0.99
    9. SitePoint1.00
    10. ②0.59
    11. So the "best" language isn't about raw power—it's about:1.00
    12. • How well Al models generate it0.98
    13. How readable/debuggable it is0.99
    14. ★1.00
    15. ★1.00
    16. • Ecosystem + libraries0.99
    17. ★1.00
    18. AIE1.00
    19. ★1.00
    20. • Speed of prototyping0.99
    21. ★1.00
    22. ★1.00
    23. ★1.00
    24. ★1.00
    25. ★1.00
    26. The closest thing to #1: Python0.98
    27. Most sources and dev consensus point to Python as the top choice.1.00
    28. Why Python wins:0.99
    29. • Extremely readable and short syntax (Al outputs are easier to verify)0.98
    30. • Massive ecosystem (Al, web, automation 'nta)0.97
    31. Works great for rapid prototyping, which . 90% of vibe coding0.97
    32. Doninnor friondl, → onninrto "otoor" Al ~utnute0.73
    33. + Ask anything0.97
    34. ChatGPT can make mistakes. Check important info. See Cookie Preferences.0.99
    35. Braintrust1.00
    36. WorkOS OpenAI0.96

  • 1:23 #5 done28 line(s)

    shot 5·sharpness 2092.4

    1. github. blicg0.83
    2. Blog1.00
    3. Changelog1.00
    4. Docs1.00
    5. Customer stories0.99
    6. Q0.98
    7. Try GitHub Copilot0.97
    8. See what's new0.99
    9. AI&ML0.96
    10. Developer skills0.99
    11. Engineering1.00
    12. Enterprise software0.99
    13. News &insights0.98
    14. Open Source0.99
    15. Security1.00
    16. Home / News & insights / Octoverse0.98
    17. Octoverse: Anew developerjoins0.98
    18. GitHub every second as Al leads0.99
    19. AIE1.00
    20. TypeScript to #10.95
    21. In this year's Octoverse, we uncover how Al, agents, and typed languages ore driving the biggest shifts in0.97
    22. ★1.00
    23. ★1.00
    24. ★1.00
    25. software development in more than a decade.1.00
    26. Octoverse 20250.99
    27. AlEngineer0.97
    28. EUROPE1.00

  • 1:41 #6 done10 line(s)

    shot 6·sharpness 2091.6

    1. "By GitHub contributor counts, August 20250.99
    2. **★0.52
    3. marks the first time TypeScript emerged as0.99
    4. AIE0.97
    5. ★1.00
    6. the most used language on GitHub,1.00
    7. surpassing Python by -42k contributors [..]."0.98
    8. – GitHub, October 2025 (emphasis added)0.98
    9. AlEngineer0.96
    10. EUROPE1.00

  • 2:18 #7 done6 line(s)

    shot 7·sharpness 1926.6

    1. Why TS/JS and Python?1.00
    2. Common and familiar1.00
    3. AIE1.00
    4. ★1.00
    5. ★1.00
    6. Engineering the future of Al0.99

  • 2:46 #8 skipped

    shot 8·duplicate of #7

  • 3:02 #9 skipped

    shot 9·duplicate of #7

  • 3:32 #10 done13 line(s)

    shot 10·sharpness 3101.3

    1. Why TS/JS and Python?0.99
    2. Common and familiar1.00
    3. Lots of frameworks, libraries, and examples1.00
    4. AIE1.00
    5. Fast to scaffold and run1.00
    6. ★0.98
    7. ★1.00
    8. ★0.99
    9. TS has typing support0.99
    10. LLMs are often good at outputting runnable code in these languages on the first try.0.99
    11. S0.71
    12. AlEngineer0.95
    13. EUROPE1.00

  • 3:43 #11 skipped

    shot 11·duplicate of #10

  • 4:27 #12 done7 line(s)

    shot 12·sharpness 1287.8

    1. ***0.65
    2. Are we even optimizing the right thing?1.00
    3. AIE1.00
    4. ★1.00
    5. A different perspective0.99
    6. S0.81
    7. Engineering the future of Al1.00

  • 4:38 #13 skipped

    shot 13·duplicate of #12

  • 5:03 #14 skipped

    shot 14·duplicate of #12

  • 5:28 #15 done12 line(s)

    shot 15·sharpness 1311.6

    1. *★0.78
    2. LLMs are fallible0.97
    3. AIE1.00
    4. ★1.00
    5. ★1.00
    6. ★1.00
    7. ★1.00
    8. They are nondeterministic and make mistakes, just like us humans0.99
    9. SS0.66
    10. LLMs arefallible0.97
    11. AlEngineer0.98
    12. EUROPE1.00

  • 5:58 #16 skipped

    shot 16·duplicate of #15

  • 6:31 #17 skipped

    shot 17·duplicate of #12

  • 6:51 #18 done14 line(s)

    shot 18·sharpness 1726.0

    1. 'MASTERFUL AND PROVOCATIVE" MUSTAFA SULEYMAN0.97
    2. Yuval Noah0.99
    3. Harari1.00
    4. **★0.66
    5. Al = Artificial Intelligence0.99
    6. Sapiens1.00
    7. AIE1.00
    8. ★1.00
    9. ★0.99
    10. Nexus1.00
    11. A Brief History of Information Networks1.00
    12. from the Stone Age to AI0.99
    13. M0.57
    14. Engineering the future of Al0.99

  • 7:09 #19 skipped

    shot 19·duplicate of #18

  • 7:49 #20 done14 line(s)

    shot 20·sharpness 1763.9

    1. 'MASTERFUL AND PROVOCATIVE" MUSTAFA SULEYMAN0.97
    2. Yuval Noah0.99
    3. Harari1.00
    4. Al = Alien Intelligence0.97
    5. Sapiens1.00
    6. AIE1.00
    7. ★0.99
    8. ★1.00
    9. Nexus1.00
    10. A Brief History of Information Networks1.00
    11. from the Stone Age to AI0.98
    12. M0.61
    13. WorkOs OpenAI0.91
    14. Braintrust1.00

  • 8:14 #21 skipped

    shot 21·duplicate of #20

  • 8:20 #22 done15 line(s)

    shot 22·sharpness 1863.8

    1. 'MASTERFUL AND PROVOCATIVE" MUSTAFA SULEYMAN0.97
    2. Yuval Noah1.00
    3. Harari1.00
    4. Al = Alien Intelligence0.98
    5. Sapiens1.00
    6. AIE1.00
    7. ★1.00
    8. ★1.00
    9. Nexus1.00
    10. Failure modes may be totally1.00
    11. unexpected for us humans1.00
    12. A Brief History of Information Networks1.00
    13. from the Stone Age to AI1.00
    14. AlEngineer0.96
    15. EUROPE1.00

  • 8:55 #23 skipped

    shot 23·duplicate of #15

Transcript

124 cues· 2,175 words· 11,953 chars

  1. 0:15 Daniel Zouk.
  2. 0:16 I'm the Rust SDK maintainer at Sentry.
  3. 0:19 And I want to tell you why I think Rust is the ideal language for vibe coding.
  4. 0:26 So the conventional wisdom on what language to use for agentic coding or vibe coding, however you refer to it, Rust is probably not one of the first things you think of.
  5. 0:41 Maybe you think,
  6. 0:44 Probably ChatGPT has a good idea what's the best agentic coding language, given that it's also an agent of some sort.
  7. 0:53 And it would tell you that there's no single number one language, but that Python is probably the top language.
  8. 1:01 And there's a strong number to, it said JavaScript and TypeScript when I asked it.
  9. 1:07 And I think that this is, at least in my experience,
  10. 1:11 pretty true, although I would flip the order because TypeScript seems to have come out as like the top choice for agentic coding lately.
  11. 1:22 And so there's even this article from GitHub that came out, I guess, late last year.
  12. 1:29 And it says that AI like they think that AI has pushed TypeScript to the number one language on GitHub by contributor counts, at least.
  13. 1:41 so they know it's typescript is the number one language and they they strongly suspect that that's because of people using it for ai assisted development but why are these languages python typescript javascript so ideal for vibe coding um at least in the this sort of conventional wisdom so first of all they're common and familiar languages um
  14. 2:11 So they're usually the languages you would learn if you were learning programming from scratch.
  15. 2:17 So they're easy for humans, and they also seem to be easy for LLMs.
  16. 2:22 There's also a lot of frameworks, libraries, and examples out there.
  17. 2:28 So that's helpful if you're building something new from scratch, of course, that you can build it on top of something.
  18. 2:35 And it's helpful for humans, it's also helpful for agents.
  19. 2:40 They're fast to scaffold and run.
  20. 2:42 They're dynamic languages.
  21. 2:45 They're interpreted, at least JavaScript and Python are TypeScript.
  22. 2:49 Maybe there's some light compilation down to JavaScript or something, but it's pretty easy just to run it and see what it does and then iterate on that.
  23. 2:59 And particularly for agents, the typing support is helpful so that the agent doesn't misuse types.
  24. 3:07 But, yeah, there's the any type that kind of undermines that a little bit in TypeScript and typed Python.
  25. 3:16 But, you know, overall...
  26. 3:18 These languages, LLMs, the models themselves are pretty good at outputting runnable code in the first try because these languages are simple and they impose few constraints.
  27. 3:34 So I think because of this fact that LLMs just seem to be good at writing them, people jump to these languages.
  28. 3:44 But I think something that a lot of people, in my experience, don't question as much is whether this is even something we want to optimize for.
  29. 3:54 The classic vibe coding languages are easy for the models to write, but is that even a good thing?
  30. 4:04 My argument is that the importance of it being easy for the model to write the language is overstated and in fact I would even say that in some cases it's a bad thing that these languages are easy for the models to write.
  31. 4:22 The dynamic and flexible nature of the languages is what makes it easy for the agent, or for the LLM, I should say, to write JavaScript, Python, TypeScript.
  32. 4:35 But this same flexibility also makes it very easy to make mistakes.
  33. 4:40 Sometimes even obvious mistakes, sometimes less obvious mistakes.
  34. 4:45 Adding typing is a helpful constraint, but that only gets you so far because it only gives you the type safety and also it's not a very strong type safety in TypeScript or Python.
  35. 4:59 And this is of course a problem because LLMs are fallible.
  36. 5:03 They will always be fallible because they are by design non-deterministic systems.
  37. 5:09 So hopefully in the future they get better at making mistakes less often, but I don't think this is something that would ever disappear entirely.
  38. 5:20 And so just like the smartest humans make mistakes and we need to guard against human error, we're also gonna need to guard against LLMs
  39. 5:30 error.
  40. 5:31 One way that folks often would do that, especially also in the conventional vibe coding languages, is adding tests.
  41. 5:40 This is a huge help, but there are a lot of problems with only relying on having tests and code review agents.
  42. 5:52 Firstly, if you don't prompt the agent skillfully, it'll often write the tests after the implementation, and then you just end up testing implementation details without actually testing the behavior properly.
  43. 6:06 Even with that test-driven development, though, tests usually can only prove incorrectness when they fail because it's impractical to test every single possible input combination.
  44. 6:21 you can't prove that every input produces the correct output in a lot of cases.
  45. 6:29 And then, of course, if LLMs are the ones generating the tests, they may make mistakes when writing those tests, and the same thing applies to coding review agents.
  46. 6:40 And then kind of more on a philosophical level, right?
  47. 6:42 We all know AI stands for artificial intelligence, but there's this book called Nexus I recently read and I can highly recommend it to anyone who hasn't read it yet.
  48. 6:54 It's from an author Yuval Noah Harari.
  49. 6:57 He's a historian and he has kind of a unique perspective on artificial intelligence.
  50. 7:02 So he's discussing

Chapters

  1. 0:00 Introduction and the speaker's background at Sentry
  2. 0:27 The current conventional wisdom for AI-assisted development
  3. 1:53 Why languages like Python and TypeScript are popular for AI
  4. 3:44 The hidden risks of prioritizing "easy-to-write" languages
  5. 6:40 Philosophical perspective: Alien intelligence and failure modes
  6. 9:28 Introduction to Rust and its strict compiler guarantees
  7. 10:53 Key safety features: Type, Null, and Concurrency safety
  8. 11:59 Demonstrating "Fearless Concurrency" with a code example
  9. 14:26 Why Rust constraints are an asset for autonomous AI agents
  10. 15:36 Conclusion and Sentry resources

Open at this second