[
  {
    "startMs": 1309,
    "text": "[music]"
  },
  {
    "startMs": 12480,
    "text": "Cool. Uh, let's get this going. Welcome"
  },
  {
    "startMs": 15599,
    "text": "everyone. Thanks for coming. Um, my name"
  },
  {
    "startMs": 18640,
    "text": "is Paul and, uh, I'm going to try to"
  },
  {
    "startMs": 21359,
    "text": "save you from the midday slump. So,"
  },
  {
    "startMs": 24080,
    "text": "let's, uh, let's see if we can do that."
  },
  {
    "startMs": 26320,
    "text": "Um, okay. So this talk, this cold open"
  },
  {
    "startMs": 29920,
    "text": "of this talk will make a lot more sense"
  },
  {
    "startMs": 31679,
    "text": "if you know uh what impeccable is. Uh"
  },
  {
    "startMs": 34640,
    "text": "but if you don't know what it is, it's a"
  },
  {
    "startMs": 35920,
    "text": "design, it's a design skill that turns"
  },
  {
    "startMs": 37600,
    "text": "your coding harness into a better"
  },
  {
    "startMs": 39360,
    "text": "designer and hopefully you into a better"
  },
  {
    "startMs": 40800,
    "text": "designer. Um and you can try it out. It"
  },
  {
    "startMs": 43600,
    "text": "works across all harnesses, cloth code,"
  },
  {
    "startMs": 45840,
    "text": "github, copilot, cursor, codex, etc. Um,"
  },
  {
    "startMs": 50160,
    "text": "and this talk is really about uh my"
  },
  {
    "startMs": 53120,
    "text": "approach to how to building this tool as"
  },
  {
    "startMs": 56399,
    "text": "opposed to the tool itself. Um, but"
  },
  {
    "startMs": 58559,
    "text": "let's get started with a pretty typical"
  },
  {
    "startMs": 60800,
    "text": "website. So, this is a intentionally"
  },
  {
    "startMs": 64000,
    "text": "very basic website and I want to give"
  },
  {
    "startMs": 66240,
    "text": "you that one as context to what I'm"
  },
  {
    "startMs": 68320,
    "text": "going to show you next. So, you know,"
  },
  {
    "startMs": 70320,
    "text": "it's it's a it's it's an okay website."
  },
  {
    "startMs": 72240,
    "text": "It's not great. It's pretty bland. Um"
  },
  {
    "startMs": 75119,
    "text": "but it at least it has some color some"
  },
  {
    "startMs": 77920,
    "text": "something. Um so now here is um the"
  },
  {
    "startMs": 82640,
    "text": "command make it bolder make the workflow"
  },
  {
    "startMs": 85439,
    "text": "section bolder with impeccable."
  },
  {
    "startMs": 88080,
    "text": "Um so this is with impeccable installed."
  },
  {
    "startMs": 91520,
    "text": "So now there are it's not perfect. There"
  },
  {
    "startMs": 95520,
    "text": "are some problems with this. for"
  },
  {
    "startMs": 96560,
    "text": "example, the um the section numbers that"
  },
  {
    "startMs": 99840,
    "text": "you have there, that's a very typical AI"
  },
  {
    "startMs": 101759,
    "text": "slop tail uh that uh that GPT loves to"
  },
  {
    "startMs": 104799,
    "text": "do and I mean Claude actually loves to"
  },
  {
    "startMs": 106240,
    "text": "do as well and"
  },
  {
    "startMs": 108720,
    "text": "it sometimes creeps in. So I wouldn't"
  },
  {
    "startMs": 110640,
    "text": "say this is perfect by any means but I"
  },
  {
    "startMs": 112640,
    "text": "think it's I would say bolder than"
  },
  {
    "startMs": 114399,
    "text": "before. Um"
  },
  {
    "startMs": 116880,
    "text": "now a lot of people install impeccable"
  },
  {
    "startMs": 119759,
    "text": "try it out uh and then they ask me well"
  },
  {
    "startMs": 122320,
    "text": "I don't know like maybe it could be"
  },
  {
    "startMs": 123439,
    "text": "placebo maybe the model is good enough"
  },
  {
    "startMs": 124799,
    "text": "in itself right uh not sure if it"
  },
  {
    "startMs": 126960,
    "text": "actually works well let's compare it to"
  },
  {
    "startMs": 130239,
    "text": "this is exactly the same project GPD55"
  },
  {
    "startMs": 132959,
    "text": "on extra high um same exact prompt"
  },
  {
    "startMs": 136640,
    "text": "without impeccable"
  },
  {
    "startMs": 140720,
    "text": "so there is a slight difference I would"
  },
  {
    "startMs": 142800,
    "text": "say but you be the judge of"
  },
  {
    "startMs": 145680,
    "text": "Um, so the before and after is pretty"
  },
  {
    "startMs": 148879,
    "text": "different and I'm going to talk about"
  },
  {
    "startMs": 151120,
    "text": "why that is and um and and what the"
  },
  {
    "startMs": 154720,
    "text": "approach is."
  },
  {
    "startMs": 156800,
    "text": "So designing at the speed of adjectives"
  },
  {
    "startMs": 159200,
    "text": "and um to get the elephant out of the"
  },
  {
    "startMs": 161680,
    "text": "room, they are not just adjectives,"
  },
  {
    "startMs": 163519,
    "text": "they're also verbs. Uh but some of them"
  },
  {
    "startMs": 166560,
    "text": "are adjectives."
  },
  {
    "startMs": 170080,
    "text": "First off, the role of the engineer and"
  },
  {
    "startMs": 172080,
    "text": "the design are blurring. Um, I've seen"
  },
  {
    "startMs": 174720,
    "text": "more people become design engineers than"
  },
  {
    "startMs": 177360,
    "text": "ever. Designers are moving into code,"
  },
  {
    "startMs": 179519,
    "text": "engineers are moving into design and"
  },
  {
    "startMs": 181440,
    "text": "vice versa. Some of the best comments I"
  },
  {
    "startMs": 183599,
    "text": "get from people who walk up to me and"
  },
  {
    "startMs": 185360,
    "text": "say like, well, impeccable gives me this"
  },
  {
    "startMs": 187920,
    "text": "shared language of design that everybody"
  },
  {
    "startMs": 189840,
    "text": "can communicate now in. But really,"
  },
  {
    "startMs": 192239,
    "text": "these worlds are all colliding in my"
  },
  {
    "startMs": 194159,
    "text": "opinion and the typical handoff is kind"
  },
  {
    "startMs": 197680,
    "text": "of broken. like the the waterfall"
  },
  {
    "startMs": 199440,
    "text": "process of I start with design, maybe I"
  },
  {
    "startMs": 202000,
    "text": "start with like the what on the PM side,"
  },
  {
    "startMs": 204720,
    "text": "then it goes into design and then"
  },
  {
    "startMs": 205920,
    "text": "there's like this handoff and this"
  },
  {
    "startMs": 207120,
    "text": "critique process and then ultimately the"
  },
  {
    "startMs": 209360,
    "text": "engineer builds it. That thing is just"
  },
  {
    "startMs": 211360,
    "text": "not real anymore. I don't know about"
  },
  {
    "startMs": 212959,
    "text": "you, but um I think that's starting to"
  },
  {
    "startMs": 215599,
    "text": "collapse very quickly in small and large"
  },
  {
    "startMs": 218080,
    "text": "companies."
  },
  {
    "startMs": 221120,
    "text": "And when it comes to the actual work of"
  },
  {
    "startMs": 223519,
    "text": "design,"
  },
  {
    "startMs": 225360,
    "text": "I feel like there's right now kind of"
  },
  {
    "startMs": 226879,
    "text": "like two worlds that you can play in."
  },
  {
    "startMs": 228879,
    "text": "Either it's direct manipulation sort of"
  },
  {
    "startMs": 231840,
    "text": "like in the pixel space, right? So"
  },
  {
    "startMs": 233519,
    "text": "you're in Figma and you're manipulating"
  },
  {
    "startMs": 235840,
    "text": "margins, paddings, whatever it is. Um or"
  },
  {
    "startMs": 238400,
    "text": "you're in web flow or you're in some"
  },
  {
    "startMs": 239760,
    "text": "other tool where you directly"
  },
  {
    "startMs": 241200,
    "text": "manipulating the final output uh or the"
  },
  {
    "startMs": 244560,
    "text": "design. Now that's fine, but then you"
  },
  {
    "startMs": 247680,
    "text": "feel like you're getting behind because"
  },
  {
    "startMs": 248959,
    "text": "everybody else is like agentic and going"
  },
  {
    "startMs": 251840,
    "text": "crazy with loops and loop maxing, right?"
  },
  {
    "startMs": 253680,
    "text": "And now you're going on the other"
  },
  {
    "startMs": 254720,
    "text": "extreme, which is simply telling your"
  },
  {
    "startMs": 256799,
    "text": "agent, please design this for me or"
  },
  {
    "startMs": 258239,
    "text": "please build this. But it feels like to"
  },
  {
    "startMs": 260560,
    "text": "me at least there's this there's this"
  },
  {
    "startMs": 262800,
    "text": "middle ground that we haven't explored"
  },
  {
    "startMs": 264479,
    "text": "enough like what is the exact level of"
  },
  {
    "startMs": 266880,
    "text": "control? How can we insert the human in"
  },
  {
    "startMs": 268800,
    "text": "the loop at the exact right time?"
  },
  {
    "startMs": 272639,
    "text": "Uh and so this this came to mind then"
  },
  {
    "startMs": 275440,
    "text": "it's pretty popular tweet a while ago."
  },
  {
    "startMs": 277680,
    "text": "Um so centering a diff with opus. Yes."
  },
  {
    "startMs": 282560,
    "text": "Uh not a great use of opus. Uh but you"
  },
  {
    "startMs": 286240,
    "text": "know it kind of speaks to that problem."
  },
  {
    "startMs": 288160,
    "text": "It's like okay well when do we use which"
  },
  {
    "startMs": 291199,
    "text": "tool? When is which tool adequate to do"
  },
  {
    "startMs": 293840,
    "text": "what job?"
  },
  {
    "startMs": 296639,
    "text": "In my opinion right now for most work"
  },
  {
    "startMs": 300240,
    "text": "the direct manipulation of padding,"
  },
  {
    "startMs": 302320,
    "text": "margins, spacing and so on is actually"
  },
  {
    "startMs": 304800,
    "text": "at a elevation that's too low."
  },
  {
    "startMs": 308960,
    "text": "And on the other hand um if you go fully"
  },
  {
    "startMs": 312880,
    "text": "agentic you get web pages like this now"
  },
  {
    "startMs": 315520,
    "text": "that you would you would get this in"
  },
  {
    "startMs": 316639,
    "text": "like 2022 I would say. So everybody here"
  },
  {
    "startMs": 319039,
    "text": "can point out the typical AI slot"
  },
  {
    "startMs": 320800,
    "text": "patterns. Um but the purple gradients"
  },
  {
    "startMs": 323919,
    "text": "were actually way beyond that. there are"
  },
  {
    "startMs": 325199,
    "text": "no purple gradients anymore uh in most"
  },
  {
    "startMs": 327840,
    "text": "frontier models. Um now we just get this"
  },
  {
    "startMs": 331039,
    "text": "and it's kind of wears a different suit"
  },
  {
    "startMs": 332560,
    "text": "as Claude would say. Um but uh but it"
  },
  {
    "startMs": 336560,
    "text": "really is slop as well. It's just a"
  },
  {
    "startMs": 338400,
    "text": "moving target. So now we have I call"
  },
  {
    "startMs": 340400,
    "text": "this claude beige. We have these"
  },
  {
    "startMs": 342000,
    "text": "instrument serif fonts. We have italics."
  },
  {
    "startMs": 344560,
    "text": "And it's not necessarily a bad design."
  },
  {
    "startMs": 346639,
    "text": "It's just everything that comes out of"
  },
  {
    "startMs": 348320,
    "text": "claw design looks that way. And when"
  },
  {
    "startMs": 351280,
    "text": "everything looks the same, you kind of"
  },
  {
    "startMs": 352560,
    "text": "get algorithmic unilo as I call it. Uh"
  },
  {
    "startMs": 356560,
    "text": "so also not great."
  },
  {
    "startMs": 358800,
    "text": "Now here's my first hot take of today's"
  },
  {
    "startMs": 361440,
    "text": "session. You cannot oneshot design."
  },
  {
    "startMs": 365199,
    "text": "Um now many people want to and maybe"
  },
  {
    "startMs": 367680,
    "text": "we'll get there eventually. I just don't"
  },
  {
    "startMs": 369440,
    "text": "think it's possible at the current"
  },
  {
    "startMs": 371520,
    "text": "moment. Uh maybe not ever. Um, I think"
  },
  {
    "startMs": 374880,
    "text": "in order to build something that really"
  },
  {
    "startMs": 377680,
    "text": "looks and feels great and is effective"
  },
  {
    "startMs": 379600,
    "text": "design, you need to have something"
  },
  {
    "startMs": 382240,
    "text": "that's context context rich which has"
  },
  {
    "startMs": 384800,
    "text": "been informed by what you're trying to"
  },
  {
    "startMs": 386880,
    "text": "do, what your audience is, what you're"
  },
  {
    "startMs": 388400,
    "text": "trying to build. And also it will have"
  },
  {
    "startMs": 390880,
    "text": "to be multi-shot. You will have to"
  },
  {
    "startMs": 392800,
    "text": "iterate on your design. Uh, and you I"
  },
  {
    "startMs": 395840,
    "text": "mean even if you're a soloreneur, but if"
  },
  {
    "startMs": 397600,
    "text": "you're not a soloreneur, you might have"
  },
  {
    "startMs": 399680,
    "text": "different people who have different"
  },
  {
    "startMs": 400639,
    "text": "opinions. Users have opinions. I mean"
  },
  {
    "startMs": 402720,
    "text": "design is messy. it's process and"
  },
  {
    "startMs": 405600,
    "text": "wishing it away and just saying like"
  },
  {
    "startMs": 407199,
    "text": "solve this problem for me agent will not"
  },
  {
    "startMs": 409680,
    "text": "unfortunately solve your problem. Um"
  },
  {
    "startMs": 414160,
    "text": "now the other problem is if you don't"
  },
  {
    "startMs": 416639,
    "text": "know the language of design or if your"
  },
  {
    "startMs": 418639,
    "text": "stakeholders don't know the language of"
  },
  {
    "startMs": 420000,
    "text": "design gets really frustrating really"
  },
  {
    "startMs": 421919,
    "text": "quickly too. So if you've ever designed"
  },
  {
    "startMs": 423360,
    "text": "something um you kind of know like this"
  },
  {
    "startMs": 425440,
    "text": "kind of experience um and then you know"
  },
  {
    "startMs": 428000,
    "text": "the inner feeling builds up of like"
  },
  {
    "startMs": 431039,
    "text": "and at some point you're like"
  },
  {
    "startMs": 435039,
    "text": "uh"
  },
  {
    "startMs": 436560,
    "text": "so really what you have to ask first is"
  },
  {
    "startMs": 438720,
    "text": "what's the emotional territory? What"
  },
  {
    "startMs": 440720,
    "text": "should this never feel like? What's"
  },
  {
    "startMs": 442720,
    "text": "what's the reference? What's the"
  },
  {
    "startMs": 444080,
    "text": "audience? These are all really important"
  },
  {
    "startMs": 446240,
    "text": "questions to figure out. Like you can't"
  },
  {
    "startMs": 447840,
    "text": "go to a human design studio even if it's"
  },
  {
    "startMs": 451919,
    "text": "the best design studio in the world. You"
  },
  {
    "startMs": 453759,
    "text": "go to design director and you say like"
  },
  {
    "startMs": 455759,
    "text": "you know what I want to design for my"
  },
  {
    "startMs": 457440,
    "text": "brand and a design dire not nods. It's"
  },
  {
    "startMs": 460400,
    "text": "like got it and then walks away. That"
  },
  {
    "startMs": 463199,
    "text": "doesn't make any sense. Like they"
  },
  {
    "startMs": 465199,
    "text": "obviously have to figure out what it is"
  },
  {
    "startMs": 466639,
    "text": "that you're trying to build and what"
  },
  {
    "startMs": 467680,
    "text": "you're trying to get out of it."
  },
  {
    "startMs": 470800,
    "text": "Importantly on this example even sort of"
  },
  {
    "startMs": 472960,
    "text": "like the 2026 version of AI slop you"
  },
  {
    "startMs": 475840,
    "text": "know quickly vibe coded page nobody"
  },
  {
    "startMs": 478720,
    "text": "decided anything here"
  },
  {
    "startMs": 481599,
    "text": "like you just oneshotted this nobody"
  },
  {
    "startMs": 483120,
    "text": "decided anything on this page this was"
  },
  {
    "startMs": 485440,
    "text": "just I mean it's maybe competent but"
  },
  {
    "startMs": 488000,
    "text": "completely empty"
  },
  {
    "startMs": 490879,
    "text": "so my uh thesis that I make here and"
  },
  {
    "startMs": 495199,
    "text": "what I've been trying to build over the"
  },
  {
    "startMs": 496960,
    "text": "last couple months is a tool that gives"
  },
  {
    "startMs": 499039,
    "text": "you just the amount of control needed to"
  },
  {
    "startMs": 502080,
    "text": "steer your agent in the right direction"
  },
  {
    "startMs": 504000,
    "text": "with adjectives and verbs."
  },
  {
    "startMs": 506800,
    "text": "And these are the types of words I'm"
  },
  {
    "startMs": 508879,
    "text": "talking about. Talking about words like"
  },
  {
    "startMs": 510800,
    "text": "make this bolder, make this quieter, uh"
  },
  {
    "startMs": 513279,
    "text": "distill this. So basically"
  },
  {
    "startMs": 514800,
    "text": "simplification, um polish this, um make"
  },
  {
    "startMs": 517919,
    "text": "it denser or harden this. Um harden for"
  },
  {
    "startMs": 521039,
    "text": "instance would mean uh make sure that"
  },
  {
    "startMs": 523200,
    "text": "this design works across the board."
  },
  {
    "startMs": 524640,
    "text": "Maybe the performance is bad on certain"
  },
  {
    "startMs": 526320,
    "text": "devices. Maybe it's not responsive. Um,"
  },
  {
    "startMs": 529040,
    "text": "and so these are these are words"
  },
  {
    "startMs": 531760,
    "text": "that we use in the world of design. And"
  },
  {
    "startMs": 535120,
    "text": "I noticed when seeing two people attempt"
  },
  {
    "startMs": 537680,
    "text": "the exact same task and one is an"
  },
  {
    "startMs": 539360,
    "text": "engineer who's never really touched"
  },
  {
    "startMs": 540880,
    "text": "design at all and the other one is a"
  },
  {
    "startMs": 542959,
    "text": "designer um there's a stark difference"
  },
  {
    "startMs": 546320,
    "text": "even when they use the same model and"
  },
  {
    "startMs": 547600,
    "text": "same harness everything in the output"
  },
  {
    "startMs": 549440,
    "text": "that they're getting based on the"
  },
  {
    "startMs": 550640,
    "text": "language that they use. So, I've been"
  },
  {
    "startMs": 554480,
    "text": "trying to put that language basically"
  },
  {
    "startMs": 557200,
    "text": "compress it into a skill uh and into a"
  },
  {
    "startMs": 559839,
    "text": "system to be able to express yourselves"
  },
  {
    "startMs": 561839,
    "text": "better. And so, for example, you could"
  },
  {
    "startMs": 564000,
    "text": "have like a baseline um over here, but"
  },
  {
    "startMs": 567360,
    "text": "then you make it bolder and the agent"
  },
  {
    "startMs": 570320,
    "text": "actually knows what bolder in your"
  },
  {
    "startMs": 571600,
    "text": "context means. You make it quieter, it"
  },
  {
    "startMs": 573680,
    "text": "knows what it means. You distill and it"
  },
  {
    "startMs": 576000,
    "text": "knows what it means."
  },
  {
    "startMs": 580320,
    "text": "Um, but then bolder how what does bolder"
  },
  {
    "startMs": 584080,
    "text": "mean in the context of your project?"
  },
  {
    "startMs": 585760,
    "text": "That's a really important question"
  },
  {
    "startMs": 587200,
    "text": "because as you could tell at the"
  },
  {
    "startMs": 588959,
    "text": "beginning of the demo uh in the cold"
  },
  {
    "startMs": 591519,
    "text": "open uh when you just ask GPT or claude"
  },
  {
    "startMs": 595200,
    "text": "whatever make this bolder it doesn't"
  },
  {
    "startMs": 596560,
    "text": "know anything about what you mean. Uh"
  },
  {
    "startMs": 598880,
    "text": "now it has its own sense of reference of"
  },
  {
    "startMs": 601279,
    "text": "like what boulder could mean but it will"
  },
  {
    "startMs": 603360,
    "text": "do whatever it wants right it will"
  },
  {
    "startMs": 605120,
    "text": "invent new colors it will invent new"
  },
  {
    "startMs": 607200,
    "text": "gradients. it will I mean it doesn't"
  },
  {
    "startMs": 609040,
    "text": "matter um what it actually means in"
  },
  {
    "startMs": 612880,
    "text": "impeccable"
  },
  {
    "startMs": 614640,
    "text": "not gradients not glass not neon but"
  },
  {
    "startMs": 617600,
    "text": "actually hierarchy scale decisive type"
  },
  {
    "startMs": 620320,
    "text": "so uh things that uh don't break your"
  },
  {
    "startMs": 623839,
    "text": "design system and adequately raise the"
  },
  {
    "startMs": 627040,
    "text": "attention that the user would put on it"
  },
  {
    "startMs": 631279,
    "text": "uh and this is an actual sentence from"
  },
  {
    "startMs": 633920,
    "text": "the the file that is loaded when you say"
  },
  {
    "startMs": 637040,
    "text": "bolder impeccable um show someone your"
  },
  {
    "startMs": 639920,
    "text": "work and say AI make this bolder made"
  },
  {
    "startMs": 642880,
    "text": "this bolder if they believe you you"
  },
  {
    "startMs": 644480,
    "text": "failed this is something that we"
  },
  {
    "startMs": 647120,
    "text": "actually instruct the agent to do for"
  },
  {
    "startMs": 649360,
    "text": "themselves and then the agent oftent"
  },
  {
    "startMs": 650800,
    "text": "times reflects and like I don't know"
  },
  {
    "startMs": 654320,
    "text": "think I did some bad work here"
  },
  {
    "startMs": 657680,
    "text": "an adjective is a light wart and a light"
  },
  {
    "startMs": 661920,
    "text": "wart such a German term and Matt Puk"
  },
  {
    "startMs": 665200,
    "text": "actually talked about this recently uh"
  },
  {
    "startMs": 666720,
    "text": "who also makes some really excellent"
  },
  {
    "startMs": 668399,
    "text": "skills. And so this is uh stolen by one"
  },
  {
    "startMs": 670959,
    "text": "of his tweets, but I really think uh um"
  },
  {
    "startMs": 674240,
    "text": "it makes a lot of sense. Um these are"
  },
  {
    "startMs": 676720,
    "text": "words that h that you basically imbue"
  },
  {
    "startMs": 679360,
    "text": "with meaning that you infuse with"
  },
  {
    "startMs": 681360,
    "text": "meaning and they mean something. They"
  },
  {
    "startMs": 683120,
    "text": "already mean something to the model, but"
  },
  {
    "startMs": 684880,
    "text": "then you translate this into the area"
  },
  {
    "startMs": 688320,
    "text": "that you're interested in"
  },
  {
    "startMs": 690880,
    "text": "and the right altitude really shifts."
  },
  {
    "startMs": 694399,
    "text": "Um, now I don't think you can solve any"
  },
  {
    "startMs": 698399,
    "text": "problems at that altitude. That's super"
  },
  {
    "startMs": 700160,
    "text": "important too. Uh, there is still room"
  },
  {
    "startMs": 702320,
    "text": "for the very beginning. For instance, if"
  },
  {
    "startMs": 704640,
    "text": "you just want exploratory work and you"
  },
  {
    "startMs": 706320,
    "text": "just want to get something on the screen"
  },
  {
    "startMs": 707839,
    "text": "or you just want the last amount of"
  },
  {
    "startMs": 710240,
    "text": "polish. I don't think right now uh AI is"
  },
  {
    "startMs": 713920,
    "text": "good enough to replace humans for the"
  },
  {
    "startMs": 715360,
    "text": "last 5, 10, maybe even 20% of the work"
  },
  {
    "startMs": 717920,
    "text": "to get it from good to great. Uh but it"
  },
  {
    "startMs": 722480,
    "text": "does actually make it better. Now I've"
  },
  {
    "startMs": 725600,
    "text": "built tools for more than 20 years. I"
  },
  {
    "startMs": 727839,
    "text": "guess I'm dating myself here. Uh but"
  },
  {
    "startMs": 730480,
    "text": "whenever I build a tool, I actually map"
  },
  {
    "startMs": 731920,
    "text": "the workflow first. I think about what"
  },
  {
    "startMs": 734000,
    "text": "is the actual workflow of the target"
  },
  {
    "startMs": 735680,
    "text": "user that I'm trying to help. And the"
  },
  {
    "startMs": 738399,
    "text": "workflow of design is not linear. It's"
  },
  {
    "startMs": 740800,
    "text": "messy. Has many different stages. Uh"
  },
  {
    "startMs": 743519,
    "text": "there are different people involved. And"
  },
  {
    "startMs": 745680,
    "text": "so I kind of mapped every part of the"
  },
  {
    "startMs": 747440,
    "text": "process and thought about well what are"
  },
  {
    "startMs": 749360,
    "text": "the injection points along that route"
  },
  {
    "startMs": 752079,
    "text": "and how can we help along that route."
  },
  {
    "startMs": 755040,
    "text": "And so you start with initialization"
  },
  {
    "startMs": 756800,
    "text": "with the shaping process of design uh"
  },
  {
    "startMs": 759279,
    "text": "you go over to uh crafting to iterating"
  },
  {
    "startMs": 762639,
    "text": "uh with all these verbs and adjectives"
  },
  {
    "startMs": 764320,
    "text": "and then to the actual harden and polish"
  },
  {
    "startMs": 766160,
    "text": "phase and finally you kind of bring it"
  },
  {
    "startMs": 768240,
    "text": "back to your design system. You do some"
  },
  {
    "startMs": 769920,
    "text": "tech debt cleanup design deck cleanup."
  },
  {
    "startMs": 773200,
    "text": "Um, now here's one example of a command"
  },
  {
    "startMs": 776320,
    "text": "that I wasn't actually sure about"
  },
  {
    "startMs": 777519,
    "text": "shipping and it's called overdrive. Um,"
  },
  {
    "startMs": 781519,
    "text": "I basically built this uh built this"
  },
  {
    "startMs": 783440,
    "text": "command in Impeccable that uh tries to"
  },
  {
    "startMs": 786320,
    "text": "create something completely over the top"
  },
  {
    "startMs": 788160,
    "text": "and it was kind of like a a half of a"
  },
  {
    "startMs": 790560,
    "text": "joke maybe. Um, but people absolutely"
  },
  {
    "startMs": 794160,
    "text": "are rabbit about this about this thing."
  },
  {
    "startMs": 795839,
    "text": "Um, and they love it. Um, and so this is"
  },
  {
    "startMs": 798079,
    "text": "just one example of I mean this is not"
  },
  {
    "startMs": 800160,
    "text": "oneshotted. Uh this is a um sort of"
  },
  {
    "startMs": 802959,
    "text": "realistic event horizon shader um that"
  },
  {
    "startMs": 805600,
    "text": "I've built for my shader library radiant"
  },
  {
    "startMs": 807760,
    "text": "shaders. But I did actually use the"
  },
  {
    "startMs": 809600,
    "text": "overdrive command for this to uh"
  },
  {
    "startMs": 812320,
    "text": "overdrive it even more uh and make it"
  },
  {
    "startMs": 814480,
    "text": "more ridiculous. So sometimes I come up"
  },
  {
    "startMs": 816639,
    "text": "with these uh commands I'm actually not"
  },
  {
    "startMs": 818720,
    "text": "sure about and I test them with the"
  },
  {
    "startMs": 819839,
    "text": "community and once it sticks uh I"
  },
  {
    "startMs": 822639,
    "text": "realize okay well there are more people"
  },
  {
    "startMs": 824079,
    "text": "who are getting joy out of this."
  },
  {
    "startMs": 826880,
    "text": "Um"
  },
  {
    "startMs": 829279,
    "text": "oh here we go. Here we go. Okay. Um, an"
  },
  {
    "startMs": 832560,
    "text": "adjective with nothing behind is just a"
  },
  {
    "startMs": 834399,
    "text": "nicer prompt. So, you really have to"
  },
  {
    "startMs": 836639,
    "text": "tell the agent what you mean by the word"
  },
  {
    "startMs": 839199,
    "text": "that you're using. Uh, it it has so many"
  },
  {
    "startMs": 841680,
    "text": "different meanings uh within its"
  },
  {
    "startMs": 844160,
    "text": "training materials that you really have"
  },
  {
    "startMs": 845360,
    "text": "to tell it. Um, now here's another"
  },
  {
    "startMs": 847600,
    "text": "thing."
  },
  {
    "startMs": 849760,
    "text": "I've gotten so many requests"
  },
  {
    "startMs": 852480,
    "text": "for an automatic way of using"
  },
  {
    "startMs": 853920,
    "text": "Impeccable. So many. It's so"
  },
  {
    "startMs": 857199,
    "text": "frustrating. Like so at least once a"
  },
  {
    "startMs": 859519,
    "text": "week somebody tweets at me say like you"
  },
  {
    "startMs": 861040,
    "text": "know what would be cool just let"
  },
  {
    "startMs": 862639,
    "text": "impeccable do all the work by itself. I"
  },
  {
    "startMs": 864560,
    "text": "don't have to decide on any command"
  },
  {
    "startMs": 865839,
    "text": "anymore. That's not the point. The point"
  },
  {
    "startMs": 867839,
    "text": "is to give you a way to steer what you"
  },
  {
    "startMs": 870959,
    "text": "want to end up with. It's never going to"
  },
  {
    "startMs": 873120,
    "text": "be a tool to oneshot design. That's not"
  },
  {
    "startMs": 875839,
    "text": "the intent. And I'll never I mean"
  },
  {
    "startMs": 877600,
    "text": "there's a pull request standing uh"
  },
  {
    "startMs": 879279,
    "text": "standing out right now that I'm going to"
  },
  {
    "startMs": 880639,
    "text": "close because of that."
  },
  {
    "startMs": 883440,
    "text": "There is no auto and there will be no"
  },
  {
    "startMs": 885120,
    "text": "auto"
  },
  {
    "startMs": 887279,
    "text": "in my opinion and uh this is a bit"
  },
  {
    "startMs": 889600,
    "text": "awkward maybe for some of the next"
  },
  {
    "startMs": 891279,
    "text": "speakers but I think taste cannot be"
  },
  {
    "startMs": 894000,
    "text": "amplified uh sorry I think taste can be"
  },
  {
    "startMs": 896639,
    "text": "amplified I think I call this amplified"
  },
  {
    "startMs": 898399,
    "text": "craft uh and I think you can work with"
  },
  {
    "startMs": 901360,
    "text": "tools to sharpen your taste and bring"
  },
  {
    "startMs": 904000,
    "text": "out more of your taste I really do not"
  },
  {
    "startMs": 905920,
    "text": "think it can be lab ground lab grown I"
  },
  {
    "startMs": 907920,
    "text": "think it is uh taste by definition sort"
  },
  {
    "startMs": 911360,
    "text": "of is uh contextual It's cultural. It's"
  },
  {
    "startMs": 915040,
    "text": "scars. And once everybody replicates the"
  },
  {
    "startMs": 917839,
    "text": "same thing, it starts to become really"
  },
  {
    "startMs": 920560,
    "text": "muddy and we don't think of it as taste"
  },
  {
    "startMs": 923040,
    "text": "anymore. That's my opinion. And I know"
  },
  {
    "startMs": 924399,
    "text": "this is like a hotly contested topic."
  },
  {
    "startMs": 926959,
    "text": "Uh, but I'm not trying to solve taste"
  },
  {
    "startMs": 928959,
    "text": "with my tools."
  },
  {
    "startMs": 930880,
    "text": "Um,"
  },
  {
    "startMs": 932560,
    "text": "that's what I got. Uh, this was about"
  },
  {
    "startMs": 934959,
    "text": "finding the right level of control for"
  },
  {
    "startMs": 936320,
    "text": "design. Thank you very much. And, uh, I"
  },
  {
    "startMs": 938399,
    "text": "think I have time for a few questions if"
  },
  {
    "startMs": 940000,
    "text": "you have any. Thank you."
  },
  {
    "startMs": 956310,
    "text": ">> [music]"
  }
]