Videos uU5Gv2h8-9g

Claude Fable, Claude Tag, and Anthropic's Culture — Cat Wu & Thariq Shihipar ft Simon Willison

index_state ready data_status ok

AI Engineer· published 2026-07-15· 0:51:30· en-US· indexed 2026-08-10 19:56

Open on YouTube

125 shot(s).

keyframes kept every frame deduplicated

What was stored

cues: 670; whisperx 670
chunks: 91; from 670 cues
keyframes: 74; kept of 125 captured
frames with text: 30; 81 lines read
chapters: 19; from the source metadata
keyframe bytes: 13.0 MB; word timings on 670 cues

Provenance

Each pipeline stage, its state and the model that produced it
stage	state	model	started	took
`fetch`	done	—	2026-08-10 18:42	2m 25s
`stt`	done	—	2026-08-10 18:45	1m 00s
`chunk`	done	—	2026-08-10 18:46	0s
`text_embed`	done	—	2026-08-10 19:55	1s
`keyframe`	done	—	2026-08-10 18:46	6m 38s
`ocr`	done	—	2026-08-10 18:52	13s
`frame_embed`	done	—	2026-08-10 19:55	13s

Frames, and what the machine read

0:02 #0 done2 line(s)

shot 0·sharpness 454.5
1. AlEngineer0.95
2. World's Fair0.97
0:03 #1 done2 line(s)

shot 1·sharpness 658.4
1. AlEngineer0.96
2. World's Fair0.99
0:10 #2 done24 line(s)

shot 2·sharpness 2720.6
1. LAB & PLATINUM SPONSORS0.99
2. Amazon AGI Lab0.98
3. ANTHROP\C1.00
4. Google DeepMind1.00
5. MINIMAX0.98
6. OpenAI0.92
7. Akamai1.00
8. arize1.00
9. aws1.00
10. Braintrust bright data0.99
11. B1.00
12. Browserbase1.00
13. docker1.00
14. :neo4j0.90
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
0:18 #3 done2 line(s)

shot 3·sharpness 256.1
1. AlEngineer0.95
2. World's Fair0.93
0:58 #4 done2 line(s)

shot 4·sharpness 285.7
1. AlEngineer0.95
2. World's Fair0.96
1:13 #5 done2 line(s)

shot 5·sharpness 269.5
1. AlEngineer0.96
2. World's Fair0.99
1:45 #6 empty

shot 6·sharpness 149.9
2:15 #7 empty

shot 7·sharpness 148.3
2:43 #8 empty

shot 8·sharpness 142.0
3:13 #9 skipped

shot 9·duplicate of #8
3:30 #10 done2 line(s)

shot 10·sharpness 269.2
1. AlEngineer0.98
2. World's Fair1.00
4:15 #11 empty

shot 11·sharpness 159.6
4:33 #12 skipped

shot 12·duplicate of #11
4:57 #13 empty

shot 13·sharpness 162.1
5:19 #14 empty

shot 14·sharpness 64.0
5:36 #15 done2 line(s)

shot 15·sharpness 224.8
1. AlEngineer0.99
2. World's Fair0.98
6:15 #16 done2 line(s)

shot 16·sharpness 218.0
1. AlEngineer0.97
2. World's Fair1.00
6:32 #17 done2 line(s)

shot 17·sharpness 278.6
1. AlEngineer0.99
2. World's Fair0.99
7:03 #18 empty

shot 18·sharpness 150.5
7:35 #19 empty

shot 19·sharpness 150.6
8:09 #20 empty

shot 20·sharpness 138.6
8:28 #21 empty

shot 21·sharpness 150.6
8:49 #22 empty

shot 22·sharpness 149.7
9:26 #23 empty

shot 23·sharpness 132.2

Transcript

670 cues· 9,350 words· 48,931 chars

chunk 0 · 0:12–0:54 · 115 words · 603 chars
0:12 Welcome to this fireside chat.
0:15 I have with me Tariq Shihipa and Kat Wu from Anthropic.
0:20 We are going to be diving deep into Claude Code, and we'll probably talk a little about this Fable thing that's been out there in the news.
0:28 Actually, on the subject of Fable, literally a minute and a half ago, Fable came back.
0:34 Like, Fable is now available to me, so if you all want to run out of the room and start using up your Fable credits, I wouldn't hold that against you.
chunk 1 · 0:42–1:21 · 110 words · 567 chars
0:42 But we're gonna have a great conversation, so please stick around.
0:45 But yeah, so please welcome Tariq and Kat for me.
0:54 Thanks for having us.
0:54 Yeah, we timed it for the chat, for sure.
0:57 Yeah, this is why it's all happening.
1:01 This year has been somewhat absurd.
1:04 It's amazing.
1:05 Claude Code came out in February of last year.
1:07 It's under a year and a half old.
1:08 And it was a bullet point on the Claude Sonnet 3.7 launch.
chunk 2 · 1:13–1:57 · 148 words · 774 chars
1:13 I'd love to hear from you.
1:14 How has what you do on a day-to-day basis changed in the past year, now that we have these coding agents that actually work for us?
1:22 I remember when we first came out with Cloud Code and Sonnet 3.7, you would give it this task, and you would have to closely monitor every single little thing that it tried to do.
1:32 I remember I would read every permission prompt extremely carefully.
1:36 I would frequently say no.
1:38 I would always say, no, no, no.
1:40 Did you check this file?
1:41 Did you check that file?
1:42 And now it's been incredible with every model generation, I feel like,
chunk 3 · 1:46–2:30 · 155 words · 769 chars
1:46 we've all gotten a chance to just take a step back, delegate a lot more of the menial implementation to Cloud.
1:53 And it's just freed up a lot of our time to think about more creative work.
1:57 Like, what is the right experience that we should be providing to our users now that we know Cloud Code can implement a lot of it?
2:04 And now with Fable, it's just a totally different step change improvement.
2:09 We see for a lot of our use cases that you can actually one-shot a ton of features with Fable now.
2:14 So it's been amazing to see the transition and to go through this with all of you in the community.
chunk 4 · 2:21–3:03 · 155 words · 770 chars
2:21 Yeah, I mean, I think I remember the first text I got about Cloud Code.
2:25 One of my best friends was like, oh, you need to go try Cloud Code.
2:28 And it was just when Opus 4 came out.
2:30 And I tried it, and I was like, oh, shit.
2:32 I need to work at Anthropic now, you know?
2:35 And that was Opus 4.
2:36 I mean, great model, but yeah, you were permission prompts.
2:39 And yeah, I think it's kind of crazy how much amnesia we have, I think, where I'm like, oh, like, auto mode has always been here, right?
2:46 Like, I don't even remember pressing, like, yes and allow.
2:50 And yeah, I think for me, the big thing that I'm trying to push myself is like, oh, we have to do higher quality work than we've ever done before.
chunk 5 · 2:57–3:38 · 130 words · 674 chars
2:57 The outputs are incredibly high quality.
3:01 I've been using it to edit videos a bunch.
3:03 And I'm like, OK, it has to meet the very exacting demands of our brand team in a couple hours, or we just can't do it.
3:10 And so yeah, I think that's how I'm trying to shift with Fable, where it's like, OK, the best work we've ever done, faster than we've ever done it before.
3:18 I've certainly been finding that myself.
3:21 Software engineering is getting harder because the level of ambition of the stuff we can take on has gone up.
3:26 I have such higher expectations of myself now that I have these tools to back me up, which is fun, but it's a lot of work.
chunk 6 · 3:34–4:19 · 115 words · 620 chars
3:34 Yeah, it's all the thinking.
3:37 It's tiring.
3:38 And so what's a piece of conventional software engineering that was true a year ago that you don't think holds anymore in this new world?

Chapters

0:00 Introductions and Claude Code overview
1:22 How coding agents have changed daily workflows
3:51 Shifting focus: Product sense over manual implementation
5:09 Why modern rewrites are now beneficial
6:37 Introducing Claude Tag and team collaboration
11:38 Prioritization and internal "dog-fooding" culture
13:06 The surprise success of remote control features
14:17 Evolving code review processes and automation
17:16 Building trust in new model generations
19:18 Optimizing for capability and user experience
21:23 Reducing system prompts for frontier models
28:05 The philosophy of tool design
30:57 Safety, security, and using Auto Mode
37:53 The human element and developer ambition
41:50 Surprising use cases for Claude (e.g., video editing)
43:35 Limitations and future design aspirations
45:09 Cultural hacks for productivity
46:42 Absurd, fun projects built with Claude
49:03 Audience Q&A