Videos pMggiOb18tc
The Golden Age of AI Engineering — Alexander Embiricos & Romain Huet & Peter Steinberger, OpenAI
Scene timeline
101 shot(s).
keyframes kept every frame deduplicated
What was stored
- cues
- 307
- whisperx 307
- chunks
- 45
- from 307 cues
- keyframes
- 84
- kept of 101 captured
- frames with text
- 81
- 3,237 lines read
- chapters
- 18
- from the source metadata
- keyframe bytes
- 13.4 MB
- word timings on 307 cues
Provenance
| stage | state | model | started | took |
|---|---|---|---|---|
fetch |
done | — | 2026-08-10 11:35 | 1m 50s |
stt |
done | — | 2026-08-10 11:37 | 31s |
chunk |
done | — | 2026-08-10 11:37 | 0s |
text_embed |
done | — | 2026-08-10 19:48 | 0s |
keyframe |
done | — | 2026-08-10 11:37 | 3m 18s |
ocr |
done | — | 2026-08-10 11:41 | 35s |
frame_embed |
done | — | 2026-08-10 19:48 | 15s |
Frames, and what the machine read
-
- AlEngineer0.95
- World's Fair0.98
-
- AIEngineer0.95
- World's Fair0.99
-
- LAB & PLATINUM SPONSORS0.99
- Amazon AGI Lab0.98
- ANTHROP\C1.00
- Google DeepMind1.00
- MINIMAX0.94
- OpenAI0.93
- Akamai1.00
- arize1.00
- aws1.00
- Braintrust bright data0.98
- B1.00
- Browserbase1.00
- docker1.00
- :neo4j0.92
- ORACLE1.00
- PayPal1.00
- qodo1.00
- reducto1.00
- Sonar1.00
- Makers of0.99
- togetherai1.00
- Unblocked1.00
- WorkOS1.00
- SonarQube1.00
-
- arize1.00
- World's Fa0.95
- Google DeepMind1.00
- World'sFa1.00
- :neo4j0.93
- World'sFa1.00
- Z.AI0.99
- World's1.00
- arld's Fai0.90
- vast.ai1.00
- World's Fa0.95
- Ref.1.00
- World's Fai0.93
- RedHat1.00
- World's Fa0.97
- mezm1.00
- API0.99
- World's Fai0.89
- Modal1.00
- World'sFai0.97
- promptql1.00
- World'sFa0.98
- THE VELOCITY ROOM0.96
- World's0.97
- orld's Fair0.90
- AlEng0.88
- jineer0.99
- com1.00
- anica0.96
- ROM's Fair0.96
- World's0.99
- WorldNDER0.99
- orld's Fair0.91
- POST0.92
- MERGE1.00
- World's0.99
- orld's Fai0.90
- RICOS1.00
- HU0.98
- Keyo0.85
- twilio0.96
- World's0.98
- orld's Fai0.94
- INNGEST1.00
- World's Fai0.94
- BAND1.00
- World's Fai0.96
- Cleric1.00
- World'sFa1.00
- A ATLASS0.92
- Snorkel0.94
- World'sFa1.00
- Z.AI0.99
- World's Fai0.97
- World'sFa0.99
- PayPal1.00
- orld's Fal0.89
- :neo4j0.91
- World'sFai0.98
- Google DeepMind1.00
- RISE PRODUCT1.00
- HEAD OF DEVELO0.97
- arize1.00
- forld'sF0.83
- redu1.00
- WorkOS0.99
- World'sFa1.00
- Amazon AGI Lab0.99
- World'sFai1.00
- nAl0.97
- Microsoft1.00
- ORACLE1.00
- World's1.00
- orid's Fai0.91
- Brai1.00
- World'sFai0.97
- OpenAl0.99
- World'sFa0.94
- MINIMAX0.99
- World'sF0.96
- Z.A1.00
- aws0.86
- Wort0.98
- togefherai0.95
- World'sFa0.99
- rld's0.88
- orld's Fa0.92
- DATADOG1.00
- Res1.00
- CODE1.00
- LanceDB1.00
- World'sFa0.95
- forld'sF0.99
- arld's Fain0.86
- cognee1.00
- BeN0.69
-
- World's Fair0.99
- ORACLE1.00
- World's Fair0.94
- arize1.00
- World's Fair0.95
- Google DeepMind1.00
- World's Fair0.97
- :neo4]0.82
- World's Fair0.97
- Z.AI0.99
- World's Fair0.94
- bright data0.99
- World's Fair0.95
- Browserbas0.97
- paper compute co.0.99
- World's Fair0.97
- extend1.00
- World's Fair0.92
- vast.ai1.00
- World's Fair0.96
- Ref.1.00
- World's Fair0.99
- RedHat1.00
- World's Fair0.99
- mezmo*0.97
- World's Fair0.97
- stigg1.00
- World's Far0.97
- World's Fair0.96
- RELAI0.90
- World's Fair0.98
- VAPI1.00
- World's Fair0.98
- Modal1.00
- World's Fair0.93
- promptql1.00
- World's Fair0.95
- THE VELOCITY ROOM0.99
- World's Fair0.96
- fiddler1.00
- World's Fair0.97
- Superconduct1.00
- Surreale0.93
- World's Fair0.97
- ZERO0.98
- World's Fair0.98
- AlEngineer0.96
- comet1.00
- World's Fair0.91
- authe0.99
- World's Fair0.93
- World's Fair0.99
- SOIOOIO0.74
- World's Fair0.95
- granica1.00
- World's Fair0.93
- dash00.97
- World's Fair0.94
- PRIOR1.00
- World's Fair0.97
- DigitalOcean1.00
- World's Fair0.99
- Vemce0.91
- World's Far0.93
- POSTMAN0.99
- World's Fair0.97
- € Composio0.91
- World's Fair0.94
- World's Fair0.96
- Modular1.00
- World's Fair0.99
- MERGE1.00
- World's Fair0.94
- AUTOMATTIC1.00
- World's Fair0.96
- Buildkite0.99
- yugabyteDB1.00
- World's Far0.94
- Zed1.00
- World's Fair0.97
- Keycard1.00
- World's Fair0.95
- Meticulous1.00
- World's Fair0.98
- World's Fair0.98
- Daytona1.00
- World's Fair0.96
- twilio0.94
- World's Fair0.99
- GRAVITEE1.00
- World's Far0.95
- PlanetScale0.99
- Temporal0.96
- World's Fair0.98
- Iw Llamalndex0.93
- World's Fair0.93
- INNGEST1.00
- World's Fair0.98
- BAND1.00
- World's Fair0.97
- Cleric1.00
- World's Fair0.97
- À ATLASSIAN0.91
- World's Fair0.95
- FACTORY1.00
- World's Fair0.96
- World's Fair0.98
- baseten1.00
- World's Fair0.97
- Snorkel0.99
- World's Fair0.99
- Z.AI0.98
- World's Fair0.99
- qodo0.97
- World's Fair0.93
- PayPal1.00
- World's Fair0.96
- Gradium1.00
- World's Fair0.95
- ANTHROPIC0.98
- mazon AGI Lab1.00
- World's Fair0.95
- Browserbase1.00
- World's Fair0.99
- :neo4]0.78
- World's Fair0.96
- Google DeepMind0.98
- World's Fair0.99
- arize1.00
- World's Fair0.94
- reducto1.00
- Worlds Fair0.91
- Microsoft1.00
- World's Fair0.95
- World's Fair0.97
- OpenAl0.94
- World's Fair0.95
- WorkOS0.93
- World's Fair0.95
- Amazon AGI!0.97
- World's Far0.95
- Microsoft0.92
- World's FFair0.92
- ORACLE1.00
- World's Fair0.95
- bright data1.00
- World's Fair0.96
- Google DeepMir0.97
- Microsoft0.99
- World's F-0.94
- docker1.00
- World's Fair0.99
- Braintrust1.00
- Worl0.78
- OpenAI0.92
- World's Fair0.98
- World's Fair0.96
- Z.AI0.96
- World's Fair0.98
- ANTHROPIC0.95
- World's Fair0.96
- World's Fair0.95
- World's Fair0.99
- aws0.98
- World's Far0.89
- to0.73
- World's Fair0.97
- CAkamai0.99
- Unblocked0.99
- Airbyte1.00
- Lightrun1.00
- World's Fair0.95
- CLOUDFLARE0.92
- Worke0.90
- DATADOG1.00
- World's Fair0.97
- Word'sFar0.86
- World sFair0.91
- World's Fair0.96
- LanceDB0.97
- World's Fair0.97
- bu1.00
- World's Fair0.94
- Ravenna1.00
- CopilotKit0.97
- World's Far0.91
- descupe0.91
- Microsoft0.98
- World's Fair0.96
- TOPK0.99
- Work0.99
- cognee1.00
- World's Fair0.96
- Be0.75
- ORC0.97
- World's Far0.93
- World's Fair0.96
- 1000.71
-
- World's Fair0.97
- Engineer1.00
- mptql0.96
- Red Hat0.93
- Z.AI0.90
- Microsoft0.99
- World's Fal0.92
- World's Fal0.92
- THE VELOCITY ROOM0.99
- mezmo*0.90
- bright dta0.61
- fiddler0.94
- O INNGEST0.91
- World's Fair0.99
- World's Fair0.98
- © BAND0.85
- POSTHAN0.83
- AUTOMATTIC1.00
- Compoin0.86
- PRIOR1.00
- SROK0.71
- Norld'sFair0.87
- Z.AI0.93
- World's Fair0.94
- Keycard1.00
- :neo4j0.96
- World's Fair0.96
- World's Fair0.97
- qodo0.97
- World's Fain0.91
- GRAMITEE0.93
- Meliculous0.95
- World's Fair1.00
- Google DeepMind1.00
- À ATLASSIAN0.96
- Amazon AGI Lab0.99
- World's!0.94
- PayPal1.00
- FACTORY1.00
- Braintrust1.00
- World'sFair1.00
- World'sFair1.00
- Micro.1.00
- World's Fair0.98
- reducto1.00
- Gradium0.93
- Dauid0.53
- OpenAI0.93
- ORACLE1.00
- World's Fair1.00
- World's Fai0.88
- bright data0.88
- Crasce )0.81
- togetherai1.00
- World's Fair0.99
- World's Fair0.95
- Z.AI0.97
- Akamal0.98
- Vorld's Fair0.98
- World's Fair0.98
- builder.io0.96
- World's Fair0.97
- DATADOG1.00
- World's Fair0.96
- Akamai1.00
- sFair1.00
- ve.ai1.00
- CopilotKit1.00
- World's Fair0.98
- Unblocked1.00
- CODERI0.93
- World'sFair0.99
- Arbye0.82
- Vorld'sFai0.97
- Opti0.97
- World's Fair0.99
- cognee1.00
- World's Fair0.98
- NCORD0.99
- World'sFai0.99
- World's Fair0.97
-
- RELAI0.98
- World'sFair0.99
- VAPI1.00
- World's Fair0.97
- Modal1.00
- World's Fair0.98
- promptql1.00
- World's Fair0.95
- THE VELOCITY ROOM0.98
- World's Fair0.98
- fiddler1.00
- World's Fair0.97
- World's Fair0.97
- ZERO0.94
- World's Fair0.92
- AlEngineer0.96
- comet1.00
- World's Fair0.98
- authe0.91
- SOIO.IO0.93
- World's Fair0.99
- granica1.00
- World's Fair0.98
- dash01.00
- World's Fair0.98
- World's Fair1.00
- World's Fair0.99
- Venuce0.97
- World's Fair0.96
- POSTMAN1.00
- World's Fair0.99
- Composio1.00
- Modular1.00
- World's Fair0.99
- MERGE1.00
- World's Fair0.96
- AUTOMATTIC1.00
- World's Fair0.99
- World's Fair0.98
- Zed1.00
- World's Fair0.98
- Keycard1.00
- World's Fair0.96
- Meticulous1.00
- Daytona1.00
- World's Fair0.94
- twilio0.99
- World's Fair0.97
- GRAVITEE1.00
- World's Fair0.97
- World's Fair0.98
- Iw Llamalndex0.90
- World's Fair0.98
- INNGEST1.00
- World's Fair0.98
- BAND1.00
- World's Fair0.98
- Cleric1.00
- World's Fair0.95
- À ATLASSIAN0.97
- World's Fair0.96
- FACTORY1.00
- baseten1.00
- World's Fair0.98
- Snorkel0.89
- World's Fair0.98
- Z.AI0.99
- World's Fair0.97
- qodo0.96
- World's Fair0.96
- PayPal1.00
- World's Fair0.99
- Gradium1.00
- World's Fair0.99
- World's Fair0.98
- Browserbase1.00
- World's Fair0.97
- :neo4j0.91
- World's Fair0.98
- Google DeepMind1.00
- World's Fair0.97
- arize1.00
- World's Fair0.99
- reducto1.00
- World's Fair0.96
- Microsoft0.96
- OpenAl0.97
- World's Fair0.97
- WorkOs0.96
- World's Fair0.97
- Amazon1.00
- Lab1.00
- World's Fair0.93
- Microsoft1.00
- World's Fair0.98
- ORACLE1.00
- World's Fair0.97
- bright data0.98
- World's Fair0.97
- World'sF0.98
- docker1.00
- World's Fair0.92
- Braintrust1.00
- OpenAl0.93
- World's Fair0.95
- World's Fair0.97
- Z.AI0.99
- World's Fair0.98
- ANTHROPIC0.99
- World's Fair0.93
- aws0.99
- World's Fair0.98
- World's Fair0.99
- d'sFair0.99
- Unblocked1.00
- WorldiaFai0.76
- Airbyte1.00
- Worl0.92
- ain0.99
- World's Fair0.99
- CLOUDFLARE0.99
- DATADOG1.00
- World's Fair0.95
- Resolve.ai1.00
- World's Fair0.98
- World's Fai0.93
- Op1.00
- LanceDB1.00
- World's Fair0.97
- 810.79
- o0.74
- World's Fair0.99
- Ravenna0.97
- rld's Fair0.96
- CopilotKit1.00
- VERIS1.00
- Wor1.00
- World's Fair0.97
- t1.00
- World's Fair0.94
- TOPK1.00
- Wo0.99
- cognee1.00
- World's Fair0.97
- NCORD1.00
- World's Fair0.97
- rid'sFa0.93
- Microsoft1.00
- ,000.60
-
- Microsoft1.00
- World's Fair0.98
- OR1.00
- AlEngineer0.96
- AIE0.83
- Vorld's Fair0.99
- 'INIMAX0.94
- Worl1.00
- Un1.00
- Akamai1.00
- AlEngineer0.97
- AIE1.00
- e.ai1.00
- Vorld's Fair0.98
- Worl1.00
-
- World's Fair0.98
- Engineer1.00
- World's Fair0.98
- World's Fain0.93
- World's Fair0.94
- THE VELOCITY ROOM1.00
- World's Fal0.89
- fiddler1.00
- comet1.00
- World's Fal0.89
- Cauthe0.84
- GEST1.00
- World's Fair0.98
- World's Fai0.87
- OSTMAN0.94
- World's Fair0.95
- PRIOR1.00
- FLOK0.89
- alkido0.87
- Fair1.00
- BAND1.00
- World's Fair0.94
- Keycard1.00
- World's Fal0.91
- Atelryx0.74
- World's Fair0.94
- Cleric1.00
- World's Fai0.99
- Meticulous1.00
- Rexa0.90
- World's Fair0.99
- World's Fair0.97
- GGRAVITEE0.93
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
307 cues· 4,464 words· 23,464 chars
- 0:13 Good morning, everyone.
- 0:14 I'm Romain.
- 0:16 Hey, everyone.
- 0:17 I'm Alexander.
- 0:19 Wow, this room is incredible.
- 0:20 There's over 7,000 AI engineers here today with us.
- 0:25 And it's not just about who's talking about this technology.
- 0:29 It's also about who's actually using it and pushing the frontier every day.
- 0:33 So we couldn't be more proud to be here with all of you today.
- 0:37 And when we were thinking about this event with Alex, we kept coming back to the World's Fair.
- 0:42 And the World's Fair made actually the future visible to everyone by building it in public.
- 0:48 Ideas that previously sounded impossible were actually suddenly there.
- 0:52 People could see them, they could walk into them, and they could even start to believe in them.
- 0:57 And honestly, this event has the exact same energy.
- 1:00 The future of engineering is not arriving from somewhere else.
- 1:04 It's really being built here by the people in this room and much faster than most expected.
- 1:09 And that's why it's a little surprising that people keep saying that engineers are going away.
- 1:13 The argument is that coding is abstracted away, and therefore, eventually, we won't need engineers.
- 1:18 Well, in fact, we think it's quite the opposite.
- 1:21 Software ate the world, and then AI ate software.
- 1:27 But now, what we're here to say is that the AI engineers are eating the world.
- 1:31 AI engineers are the people here pushing the frontier.
- 1:34 Yes.
- 1:37 And you all are figuring out how this new capability can reach everyone.
- 1:42 And there has never been a better time to be an engineer, in fact, because engineering was never about writing code.
- 1:47 Engineering has always been about solving problems for yourself and for other people as well.
- 1:53 It's about taking the latest science and combining it with design, with taste, with judgment, and most of all, imagination to make something that people can actually use.
- 2:02 And in that sense, it's not the end of engineering.
- 2:05 We think it's a return to the roots of engineering.
- 2:08 And the technology we're building on is accelerating, getting faster and faster.
- 2:14 For example, we used to ship a new model every 15 months or so, and now it's about roughly every six weeks.
- 2:21 And in case you missed it, last week we launched a preview of the 5.6 series, and we're super excited to get into all of your hands.
- 2:28 Now, building on top of all these models, the rate of product progress is relentless.
- 2:34 And as a result, I don't have to tell you, the engineering feels completely different.
- 2:40 So just to go over a couple years of what for me were successive mind-blowing experiences.
- 2:45 Obviously for a long time we've had completion, and then we went to inline prediction, and then finally we had command K where you could ask a model to make a change, but they wouldn't test the work.
- 2:55 Then models started testing the work, and now we have models taking on long, hard goals until they're done.
- 3:01 And for me, each of these phases, I remember the first time was just mind-blowing, and then obviously afterwards you just get used to it and you're trying to get your work done.
- 3:09 Yeah, in fact, I can't believe that build and test loop was not even part of the models just two years ago.
- 3:14 This was a picture of me at Dev Day 2024, and I used O1 in preview at the time to build a mini drone interface from scratch.
- 3:22 And the slightly insane part is the model could not actually run the code or verify its own work.
- 3:27 and I knew the demo would work most of the time, but surely not all of the time, so I had to cross my fingers.
- 3:32 You can kind of see here that I was pretty nervous, but hey, that's me.
- 3:36 I only do live demos, so I never know what's actually going to happen each time.
- 3:40 Luckily, it did work, and by dev day of last year in 2025, I was confident enough now that the mouse could test their own work to kind of control an entire camera system and lighting system live.
- 3:50 But yeah, we've come a long way.
- 3:51 Yeah, so we refer to Roma as the demo god.
- 3:54 And before the demo, I'll ask him, so how often does demo work?
- 3:58 And he'll be like, three times out of four.
- 3:59 And we're like, all right, good luck.
loading
Chapters
- 0:00 Introduction
- 0:13 The World's Fair analogy
- 1:10 The role of AI engineers in the future of work
- 2:14 Accelerating model development cycles
- 3:09 Evolution of build-and-test model loops
- 4:03 Scaling engineering capabilities through agents
- 5:30 Defining the desired AI engineering product experience
- 7:59 The design philosophy of the Codex app
- 9:44 The open-source stack and building with API primitives
- 11:43 Expanding the ecosystem with Apps Server and plugins
- 14:20 Optimizing for "Value Maxing": Cost and Intelligence
- 15:48 Achieving high-speed inference for real-time workflows
- 16:51 Future outlook: Removing the local/cloud distinction
- 18:16 Special guest introduction: Peter Steinberger
- 18:56 Shifting from manual orchestration to managing agents
- 20:02 Three key changes for scalable agent loops
- 21:19 Redefining the bottleneck as human attention
- 22:08 Workflow example: Automating open-source issue resolution