Brandon Jacoby - Seeing Taste vs. Creating Taste as a designer

2026-04-21 - 53 min - source - Read full transcript

Rid (host)Brandon Jacoby

Key insights

Taste is the ability to know when to break rules and when to follow them, not a fixed preference for novelty.

Brandon defines tastemakers as people who know when to push beyond existing boundaries and when to quiet down and go with the flow. He argues this back-and-forth judgment, not raw creativity, is the actual skill - and it's the one thing current AI tools cannot supply on their own.

taste-and-design-judgment

Great design has always been agnostic of the tools used to make it.

Brandon pushes back on both the 'AI is deleting Figma' and 'AI can't touch human taste' extremes, arguing design has always been about decisiveness and judgment regardless of which tool happens to be popular at a given moment.

taste-and-design-judgment

The biggest unlock from AI for Brandon's client work is being able to build his own bespoke design tools.

Rather than just using existing AI features, Brandon builds custom tools on demand for whatever a project needs, calling it the most creatively empowering feeling of his career. He frames this with a John Lasseter quote about Toy Story: 'the technology inspired the art, the art challenged the technology.'

ai-as-creative-amplifier

People who already ran through walls before AI existed get the most amplification from it now.

Brandon says the trait separating highly amplified designers from capped ones predates AI: not taking no for an answer, questioning why something is a requirement, and finding workarounds when stuck. AI removes technical barriers for people who already had that initiative; it doesn't instill the initiative itself.

ai-as-creative-amplifier

AI-generated 'good enough' work still fails on visual design fundamentals that shape how users feel.

Brandon argues that things like alignment and basic composition principles directly affect whether a user feels at peace using a product, and that vibe-coded outputs often skip these fundamentals even when they look superficially polished.

ai-as-creative-amplifier

The exploratory, uncertain period after leaving a big job can be more fulfilling than having a defined role.

After leaving X, Brandon spent three deliberate months incubating side projects and helping friends' startups with no fixed direction. He found that open-ended exploration itself, not any single outcome, was the most fulfilling work he'd done in years, and it became the catalyst for starting his solo practice.

independent-practice

A solo practice trades organizational overhead for the freedom to fully be yourself as a designer.

Brandon says there is far less he has to tolerate outside of pure design work in his independent practice than in a big company; the tradeoff is picking up 'annoying side stuff' like project management and taxes himself, which he accepts in exchange for depth on work he cares about.

independent-practice

In a world where anyone can build 'good enough' cheaply, taste becomes the product being sold.

Brandon compares software to a housing market where the cost of materials and labor collapses: even if building becomes trivial, people still have to choose where to live and how a place makes them feel. He argues the same holds for software - human judgment about what to build and how it should feel remains the scarce, sellable thing.

independent-practice

Finding people you enjoy working with matters more than the company's stage when going zero-to-one.

Brandon's top advice for designers entering early-stage, zero-to-one roles is to prioritize the team fit over the opportunity itself, since in a tiny company the designer collaborates almost exclusively with founders and leadership rather than other designers.

zero-to-one-design

Starting with a bad first draft is the only way to reach a great final one.

Brandon says he now expects his first attempt at anything to look like 'absolute garbage,' and that producing visible bad ideas actually accelerates a founder's own thinking by giving them something concrete to react against, rather than trying to think a solution into existence before touching the canvas.

zero-to-one-design

The discovery that happens between AI prompts, not the prompts themselves, is where the best design work emerges.

Brandon distinguishes hands-on canvas tinkering from prompt-and-wait AI workflows: the accidental 'weird C thing' discovered while manually iterating is where real breakthroughs happen, and he says that in-between exploration is exactly what gets lost when a founder tries to vibe-code a product without a designer.

decisiveness-and-discernment

Decisiveness is becoming a defining professional trait as AI increases the number of micro-decisions in design work.

Brandon points to his time at X, where constant first-principles questioning forced him to unlearn the reflexive constraints designers usually carry. He predicts decisiveness will matter more as AI creates more raw material and options for designers to sort through.

decisiveness-and-discernment

Media referenced

TBPN - podcast - Jordi, former CEO at Party Round/Capital and one of Brandon's early mentors, now hosts this podcast.

Companies

X (Twitter) - Brandon's most recent full-time role; source of the Elon Musk first-principles design-review story and the Nikita onboarding story.
Cash App - Brandon led growth-team design there for four years, working closely with mentor Owen Jennings; formative onboarding and AB-testing experience.
Capital (Party Round) - Brandon was the first design hire; Party Round's design work is cited by the host as a standout early example of Brandon's craft.
Block - Parent company where Owen Jennings, Brandon's mentor, is a business executive.
AngelList - Client of Brandon's new independent practice; hired him to build a net-new product from the ground up.
Coinbase - Brandon embedded with a small team there late last year to shape a new, still-forming product initiative.
Figma - Brandon's primary daily design tool; he contrasts hands-on canvas tinkering with prompting AI tools.
Framer - Show sponsor; segment on its new Wireframer and Workshop AI features.
Jitter - Show sponsor; segment on its new image-to-video AI feature.
Inflight - Show sponsor (host's own product); a feedback tool for sharing prototypes and video walkthroughs.
Apple - Cited as the ultimate example of taste and high-quality, accessible design at massive scale.
Uber - Cited as an example of a software invention that made a previously inaccessible experience available to the masses - Brandon's definition of taste as craft plus accessibility.

Techniques and frameworks

Seeing taste vs. creating taste - Brandon's core framework: everyone can recognize good design, but creating taste requires knowing when to follow existing patterns and when to break them.
Craft meets consequence - Brandon's idea that deeply crafted design and commercially consequential, metric-moving work are not mutually exclusive and should reinforce each other.
Zooming in and out - The discipline of alternating between deep, exploratory iteration on a problem and knowing when to stop pushing and accept the status quo.
First-principles questioning - The Elon Musk-style practice of questioning every requirement and asking why it exists, rather than accepting inherited constraints.

Summary

Brandon Jacoby - who led growth design at early Cash App, was the first design hire at Capital/Party Round, and most recently worked at X - joins Dive Club to talk about launching an independent design practice and what separates seeing good taste from creating it. He opens with stories from X, where Elon Musk's first-principles culture forced him to question every design constraint he'd previously taken for granted, and where incoming head of product Nikita revealed a wealth of unspoken onboarding pattern knowledge that reshaped how Brandon thinks about growth design.

The episode's core argument is Brandon's distinction between seeing taste and creating it. Anyone can recognize when something is beautifully crafted, he says, but tastemakers possess a subtler skill: knowing when to follow existing patterns and when to break them, and knowing when to keep pushing a prompt or iteration versus when it's fine to leave something as-is. He frames this "zooming in and out" as the exact judgment call that current AI tools cannot make on their own, which is why he doesn't buy either extreme in the AI-and-design debate - that AI is making human craft obsolete, or that it can't touch design at all. Great design, he argues, has always been agnostic to whichever tools happen to be in fashion.

On AI specifically, Brandon describes his biggest unlock as being able to build custom design tools on demand rather than only using off-the-shelf ones, quoting Pixar's John Lasseter ("the technology inspired the art, the art challenged the technology") to describe the feeling of no longer being limited by what he can technically build. But he's clear that AI amplifies existing initiative rather than creating it: designers who already had a habit of "running through walls" and refusing to accept blockers get the most out of AI, while those who previously externalized problems don't automatically become more capable just because the tooling improved.

The back half of the conversation covers Brandon's transition to solo practice after a deliberate three-month period of open-ended exploration following his exit from X, which he found unexpectedly more fulfilling than having a defined role. He introduces "craft meets consequence" - the idea that deeply crafted work and commercially consequential, metric-moving work aren't opposed, and that the best products hold both simultaneously. He also argues that as the cost of building software collapses toward zero, taste itself becomes the scarce, sellable asset, drawing an analogy to housing: even if construction were free, people would still choose where to live based on how a place makes them feel.

Closing out, Brandon gives tactical advice for designers entering zero-to-one environments - prioritize finding people you enjoy working with over the company's stage or size - and stresses that producing bad first drafts quickly is what actually accelerates both his own thinking and a founder's. He credits mentors like Owen Jennings (Block) and Jordi (former Party Round CEO, now TBPN host) for shaping him as much as any single design crit. He closes on his hope for the practice: to bring taste that can "fight the slop" in a world of AI-abundant, easily produced software.

Notable Quotes

"I think the tastemakers know when to break the rules and when to push beyond the boundaries of what exists today and when to quiet and to go with the flow." - Brandon Jacoby

"The technology inspired the art, the art challenged the technology." - Brandon Jacoby, quoting John Lasseter

"In a world where it's super easy to build like good enough, the thing that I'm selling is how to be great." - Brandon Jacoby

"The first version of everything that I create is so bad... but starting is not where you finish and doesn't have to be where you finish." - Brandon Jacoby

"I hope for is that in a world of a lot more software and abundance and choice, there's still something great... fight the slop." - Brandon Jacoby