# 015 The Rise of AI Builders ![[images/015-Shane-Ward-episode-art.jpg]] Part of [[Pixels & Priorities]] (listed in [[Pixels & Priorities - Audio Episodes 015-031]]). Previous: [[014 When Product and Design Roles Blur]]. Next: [[016 Networking in the AI Job Market]] (Shane Ward part 2). - Show: Pixels & Priorities, hosted by Metsy Rose and J Schuh (RSS author field) - Episode: 015, published 2026-07-07 (RSS pubDate Tue 05:00 -0500); guest Shane Ward, part 1 of 2 - Audio: https://dts.podtrac.com/redirect.mp3/mcdn.podbean.com/mf/web/fkpr87kqkh7akdhr/PP_015_Shane_Part_1_Editedbmotg.mp3 - Podbean page: https://pixelsandpriorities.podbean.com/e/015-the-rise-of-ai-builders/ - Length: 44:25 (2,665.2 s measured from the mp3). The RSS `itunes:duration` says 3053 s (50:53), which does not match the file; the publisher transcript also ends at 44:20. - RSS description: "Metsy and J sit down with Shane Ward, Manager of Product Design at Storable, to explore the rise of AI builder roles. Who owns quality? How do teams make decisions when everyone can generate more ideas? And what happens if AI becomes a shortcut around experience instead of a support for it?" - Artwork: the feed's episode image (Shane Ward guest card), saved as `images/015-Shane-Ward-episode-art.jpg`. - Transcript: faster-whisper (medium, int8, CPU) on the mp3 with VAD on, beam size 5 (about 6.9 minutes of transcription). That pass skipped or garbled seven stretches (2:30-3:00, 3:02-3:28, 10:56-11:20, 17:17-17:46, 30:41-31:00, 33:34-34:05 and 34:07-34:35), which were re-transcribed in 13 short chunks with VAD off, for about 7,200 words in total. The feed's publisher transcript (Podbean VTT, `PP_015_Shane_Part_1_Editedajdzt.vtt`, about 8,200 words) was used for the outline timestamps and to check names and wording. All quotes come from the Whisper transcript (case and stray-character fixes only). Speakers are not diarized; attribution follows content and the speakers addressing each other by name (J: the intro-to-UX class, the baby analogy, his adult kids, knowledge versus understanding, the value of leadership, the layoff ripple effect, Brian Sullivan and the French Revolution; Shane: Storable, AI builders, quality, the author-to-publisher analogy, AI as decision-maker, people leadership; Metsy: firefighting, HiPPO, AI as a source of truth, RVs and tiny homes). Where a line is uncertain it says "likely." Correct spellings are **Metsy Rose** and **J Schuh** (Whisper heard "Jay"). ## Hosts - **Metsy Rose** - co-host; product manager and strategist (show open). Says every place she has worked has been trying to get out of firefighting mode; one of her teams is trying to re-stand up continuous product delivery. Has no kids, by choice. Is "obsessed" with RVs and tiny homes, which she calls her doom-scrolling relief after work. - **J Schuh** - co-host; design strategist, educator, and creative mind (show open). Teaches an intro to UX class. His kids are now adults in college. Says he has a psychology background and has studied why people do what they do since his 20s; Brian Sullivan (episodes 012-013) often comes to him asking why people behave as they do. ## Guest - **Shane Ward** - Manager of Product Design at Storable (RSS and show intro), "driving both design vision and product strategy." Says Storable is trying to become an AI-first company and leadership has given his team the tools to "wear many hats." A people leader who sees mentoring employees as a big part of his job; says he is still learning product and back-end engineering. Parent of a teenager. Has worked with the hosts before ("We get to do some work sometimes"). Metsy notes Storable's products serve communities such as RV and tiny-house communities. Part 2 is episode 016 Networking in the AI Job Market. ## Outline (timestamps are mm:ss) - **0:00 Show open.** Metsy Rose introduces herself and J Schuh and welcomes Shane Ward, Manager of Product Design at Storable. Since the hosts have been discussing how design and product roles will look in the future and AI's impact on them, she asks Shane, who works on both sides, how he sees these roles. - **1:02 An AI-first company.** Shane says Storable recently began trying to become more of an AI-first company, an "awakening" that they have to change how they work. Its AI products are slim so far, but leadership has empowered people to wear many hats and given them the tools they need, an investment he says means the world to him. - **1:47 How fast can we build?** His team has run tests of how fast they can build things and what AI can and can't do. He cites a friend's quip that you almost have to be unemployed to keep up with AI, studying about 12 hours a day. They are defining new processes so AI doesn't slow them down, since it is easy to fall into optimizing endlessly and never building. - **2:39 AI-built versus hand-built.** In one test they built something quickly with AI and then built it themselves; the gap between the two was small because they had to correct so much of the AI output, though he says that may have been their use case and AI will improve. - **2:57 AI builders.** Shane describes the term "AI builders": design, product and engineering are told to just go build and let AI fill the gaps. He finds it exciting but a little scary, since he is still learning product and back-end engineering and "wouldn't even know if it was wrong." AI has leveled the playing field for the first level of knowledge across roles. - **3:39 Who owns the experience?** J (who teaches an intro to UX class) says product and design are both being asked to lean into AI, which creates role ambiguity. Product traditionally finds the features and value that meet business goals; design makes an experience people want to use and don't delete from their phone. His students almost universally delete an app that doesn't match its App Store description or their mental model, regardless of the hours of analysis and design behind it. - **5:41 Merging roles.** J welcomes product people using Figma Make and designers getting more involved in the business, a chance to counter the myth that UX equals UI. But who owns it? Today product-design stalemates are settled by bringing in UX research and usability testing; when roles cross over, who decides, when getting it wrong means people delete what you made? - **7:34 Quality in the flood.** Shane says with AI there will be 100 apps next to yours and perhaps 80% get deleted, so staying on top of quality becomes critical: quality in work, in decisions, and in inviting diverse perspectives. Building in a silo means building for yourself; without user interviews and empathy, products all look the same, following "industry standards" just to avoid deletion. - **8:39 Too many good choices.** The hard part is not creating things or choosing between them but sifting the massive output for the quality bits. Teams will get smaller and need a lead; if titles are stripped and everyone is a builder, perhaps it is the most senior builder, or leadership decides, but then how does leadership connect with users and the data? "It depends" is the usual answer. - **9:58 Time for polish.** Shane hopes speed unlocks the polish teams never get to, since work is always MVP: cool interactions, better design, and features aimed at what's possible tomorrow. In the end it is a business decision. He asks Metsy, the product owner on the call, where product sits. - **10:57 Beyond firefighting.** Metsy hopes AI efficiency moves teams from one MVP to the next toward roadmaps that actually get worked instead of presented and shelved. If design and product can't partner to move beyond firefighting, AI becomes another firefighting tool. One of her teams is trying to re-stand up continuous product delivery. She won't say AI can't innovate, but for now innovation is tied to what humans want, so humans need to meet that need. - **13:13 The baby analogy.** J compares it to a struggling couple who decide a baby will fix their relationship: the communication problems between design and product will only be magnified by AI. And AI is in its infancy; if the "parents" don't get their act together, what happens when it reaches its teenage years? - **15:00 From author to publisher.** Shane, a parent of a teenager, wants an age of creativity where ideas get out fast to see if they flop. He has done endless documentation, people management, pixel pushing and mobile and tablet versions, and wants that to subside so he can move upstream into strategy and research, finding pain points he feels too. Production will be taken over at some point; the question is whether it will be good. He told his team they are moving from independent authors writing as many books as they can to publishers with many authors, who must think like the business about hits. - **17:16 Who rises.** Those most creative at working within a system will bubble up, and people with no day job in the way will be enabled to dream something up and put it out there. - **17:41 Divorces at seven or eight.** J, whose kids are now adults in college, says he saw many parents divorce when their kids were around seven or eight: the baby didn't fix the problems. He wonders whether companies that thought AI would fix everything, with foundations not where they should be, will see similar "divorces." Shane asks who the parents are: business and employee, design, product and engineering, or business and customer; it could be all of them. - **19:10 Decisions not made in a vacuum.** Metsy, who has no kids by choice, is most concerned about decisions that are reactive or made in a vacuum. Teams need a paradigm for how decisions get made; instead of the HiPPO (highest paid person's opinion) without explanation, she hopes for transparency on why decisions are made and the information behind them (for example, why a new AI platform was picked). - **20:56 Helping each other through change.** The rate of change will require everyone to communicate more. Some people handle frequent new information easily; others don't, so teams must help each other move in the same direction. If the people in the room can't figure out how to use new technology together, struggles remain. - **21:57 Knowledge is not understanding.** J recalls earlier episodes' I, T, pi and comb shapes and unicorns ([[008 Unicorns, T-Shapes, and the Myth of Mastery]]), and says he read that knowledge does not equal understanding. AI output is knowledge; understanding comes from experience, working with things and seeing what works, which ChatGPT or Copilot can't replicate. A weekend on ChatGPT doesn't make him a product manager, and leaders who come back from an article or conference saying all you need to do is this ignore the teams that must get on board to execute. - **24:55 AI as decision-maker.** Shane raises what happens when AI makes the decisions: push to a small percentage, then go or no-go from data. That removes experience, perspective and the human; it might work better in a model that blasts output at the market until enough sticks, but few people would be happy with an "AI boss" deciding whether their work gets used. - **25:51 Not a source of truth.** Metsy says the dangers of giving AI too much latitude as a source of truth are already visible: she cites recent articles about companies whose internal AI tools for metrics and sales data were found months later to be hallucinating the numbers. Some tools seem to want to please you with a quick answer. She repeats the human-AI-human pattern: give good input, then review the output against your own experience, because humans will be blamed. AI is built on human data, so it carries human mistakes and bias, and it is "not a reason to outsource our own brains." - **27:50 Checks and balances.** She warns against leaders who go to a conference and decide to lay off half their staff because AI will figure it out, releasing people into an already fragile hiring market. - **28:30 Seeing beyond the radar.** J (likely) describes forecasting as watching moving points on a radar; he tends to see things outside the radar heading toward it, and by the time they hit it, it's too late. The bigger the ship, the harder it is to steer; smaller, agile companies with tiny teams may emerge, as Apple started with a couple of guys in a garage, a hopeful "Star Trek future." - **30:03 What is leadership worth?** J asks what the value of leadership is. Big organizations have a highly compensated layer of leaders; if AI can take over vision and enabling work, a company could save millions with one person at the top getting information from AI and deep subject-matter experts working with AI beneath. Leaders talk about cutting middle management and lower-paid roles, but as the AI "baby" grows smarter and more independent, are they considering their own jobs? - **31:49 The end game.** Shane thinks many companies want no employees, since employees are the biggest expense, and to blast out work until some sticks. As a people leader, much of his value is mentoring; with few employees there is nowhere to put that experience, so he would have to become a doer: "If you can do it and you can sell, you're good." He'd rather fill a company with people each doing 1000x, which would dominate a market, but thinks leadership wants margin and to own a niche. - **33:49 Who buys if no one has a job?** Metsy finds the no-employee worldview disturbing: who are you selling to if no one has a job? Any extreme future has a negative end if it ignores the human toll; she calls for measured, thoughtful leadership in constant collaboration with multiple perspectives rather than going to a conference and coming home to fire 1,000 people. She would love AI focused on helping people be better at what they're skilled and built to do, offloading work so talent shifts to harder, more innovative problems. - **36:27 The layoff ripple effect.** J says laid-off people stop spending, and employed people who see layoffs stop spending too; money flowing through the economy is its lifeblood and confidence drives the stock market. Falling sales prompt more layoffs, which makes no sense. He suggests typing the question into AI: what is the long-term impact of everyone laying people off? Metsy prefers Shane's idea of amazing, ultra-productive people; Shane says the balance of leaders who think that way is encouraging. - **38:43 A boom in new companies.** Shane says the people released into unemployment will spin up concepts, and we could see the biggest boom in new companies in anyone's generation. - **39:00 Nothing to lose.** J says Brian Sullivan, a very logical person, often asks him (because of his psychology background) why people do irrational things; J's answer is that you can't expect rational behavior from irrational people. Some laid-off people will start businesses, but others will be angry and ask what harm they can do with these tools; the French Revolution didn't happen for no reason. Society must give people purpose and a way to provide for their families. - **41:21 Creativity born of necessity.** Shane prefers the positive outcome: people get creative to make a paycheck, there is so much good in people, and AI is unlocking great ideas that few ever had time for. Product designers should keep learning where it breaks and how to fix it. - **42:24 RVs, tiny homes and tech literacy.** Metsy wants to live in Shane's version. She mentions that some of Storable's products serve RV and tiny-house communities, which she loves watching, and hopes for broader technology understanding in the general population, since many hear about the fears of technology without exposure to it. - **43:39 Community.** "We get past this with community," Metsy says; J agrees, calling it a great way to end the first episode with Shane. Metsy thanks Shane, and the discussion continues next episode. Show outro. ## Standout quotes (verbatim from the Whisper transcript) - 2:09 "You almost have to be unemployed to keep up with AI right now" (Shane, citing a quote a friend mentioned) - 2:59 "There's a term floating around called AI Builders now." (Shane) - 3:23 "I wouldn't even know if it was wrong" (Shane, on AI-built back-end code) - 3:32 "AI has really leveled the playing field for that first level of knowledge across all of these, these roles." (Shane) - 7:28 "if we get it wrong, people are going to want to delete what we create." (J) - 8:06 "If I build in a silo, I'm building for myself." (Shane) - 8:24 "then all the products start looking the same because we just do air quotes industry standards" (Shane) - 10:06 "Like we never get to work on the polish, right?" (Shane) - 12:01 "I'm just hopeful that AI will help us find ways to get efficiency in the right places to release us to spend time on the human work of innovation." (Metsy) - 13:35 "not really communicating well decides that a baby is gonna fix it all" (J) - 18:09 "I saw a lot of divorces around seven, eight years old" (J) - 20:00 "If we continue to have HIPPO happen, the highest paid person's opinion" (Metsy) - 22:11 "You can watch a video. You can read a book. You can gain knowledge. However, to create results, you have to have understanding." (J) - 23:20 "I got your job because I was on chat GPT all weekend." (J, playing a would-be product manager) - 25:44 "And then our AI boss is telling us if we get to use it or not." (Shane) - 27:37 "it is still ultimately a set of tools and useful technology and not a reason to outsource our own brains." (Metsy) - 30:07 "What is the value of leadership?" (J) - 32:59 "If you can do it, and you can sell, you're good." (Shane) - 34:13 "Who are you selling your products to if no one has a job?" (Metsy) - 35:08 "I went to a conference and I came home and I fired a thousand people because surely it's not that hard, right?" (Metsy, describing decision-making she fears) - 36:59 "money flowing through our economy is it's lifeblood, right?" (J) - 40:17 "History has shown the French Revolution didn't occur for no reason" (J) - 43:39 "We get past this with community" (Metsy) ## Books mentioned - No books are named in this episode. (J says "you can read a book" and "I read that knowledge does not equal understanding" without naming a source.) ## Things mentioned - People and companies: Shane Ward, Storable, Brian Sullivan (episodes 012-013), Apple (founded in a garage). - Tools and media: Figma, Figma Make, ChatGPT, Copilot, the App Store, Star Trek. - Ideas: AI-first companies, "AI builders," optimizing instead of building, AI leveling the first level of knowledge, UX equals UI as a myth, usability testing to settle product-design stalemates, quality and perspective diversity, building in a silo, the author-to-publisher shift, the baby analogy for AI, HiPPO (highest paid person's opinion), knowledge versus understanding, human-AI-human review, AI hallucinating business metrics, AI as decision-maker, the value of leadership, the no-employee end game, layoff ripple effects on spending, a boom in new companies, the French Revolution, continuous product delivery, technology literacy, community. - Callbacks: I, T, pi and comb shapes and the unicorn from [[008 Unicorns, T-Shapes, and the Myth of Mastery]]; "human, AI, human" from earlier episodes; Brian Sullivan from [[012 UX Research to Product Strategy in the AI Era]] and [[013 Build Your Network Before You Need It]]; the Shane Ward preview at the end of [[014 When Product and Design Roles Blur]]. ## Good work, productivity and career ideas - **Test AI against the hand-built version.** Shane's team built the same thing with AI and by hand and found the gap small once AI output was corrected; measure before assuming speed (1:47-2:57). - **Don't optimize instead of building.** Define processes so AI speeds you up rather than trapping you in endless optimization (2:23-2:39). - **Know where you can't judge the output.** If you couldn't tell whether AI built something wrong (for Shane, back-end engineering), learn that area or bring in someone who can (2:57-3:39). - **Settle stalemates with users.** When product and design disagree, bring in research and usability testing rather than titles (6:45-7:34). - **Guard quality and perspective.** With far more output, the skill is curating the quality bits and inviting diverse perspectives so you don't build for yourself (7:34-9:58). - **Think like a publisher.** When you can produce many "books," manage the portfolio for hits instead of producing as many as possible (16:28-17:16). - **Use AI time for polish.** If production gets faster, spend the saved time on interactions and tomorrow's features, not just more MVPs (9:58-10:38). - **Fix the relationship before the "baby."** AI magnifies existing communication problems between product and design; repair them first (13:13-14:58, 17:41-19:06). - **Ask why decisions were made.** Push for transparency on the reasoning and data behind decisions instead of HiPPO (19:10-21:54). - **Help slower adopters.** People absorb change at different rates; communicate more and bring the team along together (21:04-21:54). - **Knowledge is not understanding.** Pair AI-generated knowledge with real experience before acting on it (21:57-24:52). - **Human, AI, human.** Give good input, then check AI output against your experience; you will be the one blamed for it (26:02-27:50). - **Look beyond the radar.** Pull back for a bigger view of what's heading your way before it becomes urgent (28:30-30:01). - **Become a doer who can sell.** As roles shrink, Shane advises being an expert who can do the work and sell it (32:46-33:10). ## Claims check | Claim | Check | |---|---| | HiPPO stands for "highest paid person's opinion" (20:00) | Correct; the term is usually credited to web-analytics writer Avinash Kaushik. | | Companies' internal AI tools were found months later to be hallucinating sales numbers (26:11) | Metsy cites "recent articles" without naming them; not verified here. Hallucinated figures from generative AI tools are a well-documented risk. | | Apple was created in a garage by a couple of guys (29:35) | Apple was founded in 1976 by Steve Jobs, Steve Wozniak and Ronald Wayne, with early work in the Jobs family garage in Los Altos; Wozniak has said the garage's role is somewhat mythologized. | | Many parents divorce when their kids are about seven or eight (18:08) | J's personal observation. U.S. Census Bureau reports put the median length of first marriages that end in divorce at roughly eight years. | | The French Revolution "didn't occur for no reason" (40:17) | Historians tie it to fiscal crisis, food shortages and high bread prices, and social inequality, among other causes. | | Episode length | RSS 50:53 (3053 s); measured mp3 44:25 (2,665 s). | ## Added context - Audio-only RSS episode (not on the YouTube channel). Part 1 of 2 with Shane Ward, previewed at the end of episode 014; part 2 is 016 Networking in the AI Job Market.