All Episodes

Why your AI-designed brand is invisible, and the 3-step fix

with Jacqueline Schmidt · Radiance

September 8, 202601:01:32New York, NY

Why your AI-designed brand is invisible, and the 3-step fix

0:000:00

Show Notes

Jacqueline Schmidt grew up two subway stops from tigers.

Funding stage: Series A Plus. Radiance is a business unit of Infinity Constellation, the AI-native company builder that announced a $24M Series A on June 9, 2026, with participants including Freestyle Capital, Backed VC, Rafferty Holdings, Isomer, Oxford Funds, BY Ventures, Gaingels, Millennia Capital, Charlie Songhurst, and Henry Ford III. Radiance has no announced standalone round, so the stage travels through the parent: backed through Infinity Constellation's Series A, not a Radiance raise on its own.

The Bronx Zoo was her backyard. Her favorite exhibit was the sloth, which she went back to visit twenty-five years later and found out had died. She spent summers collecting salamanders in the woods and carrying them home in her Smurf boots, fifty at a time, to the ongoing horror of her grandparents. She drew animals constantly. Then she sold the drawings. Then the drawings became a company.

Today she is CEO of Radiance, an AI-powered design agency running a proprietary Creative OS called The Source®, with clients including Coca-Cola, MoMA, CBS, NBC, Guess, Crate & Barrel, Barnes & Noble, Anthropologie, West Elm, and Siggi's.

The through-line from salamanders to Series A is the thing this episode is actually about: what survives when the tools stop being the advantage.

Because right now, they have stopped. Jack's Instagram feed is wall-to-wall ads promising a design agency can go from $10K a month to $100K a month by plugging in one AI tool. Then she takes real meetings and gets handed real decks. Same fonts. Same colors. Same soulless template with a different logo on top. “The originality is kind of lost,” she says. The tools got more powerful and taste got more rare, so everyone is swinging the same hammer and almost nobody is building anything memorable.

She has a name for what happens next. Digital Darwinism. It is the title of the TEDx talk she gives at TEDxSouthlake on November 8, 2026, and the argument is blunt. In a landscape where every founder has the same generative tools, the selection pressure is no longer access. It is memorability. Who evolves. Who is sticky. Who is left. She frames it as a question you should be able to answer in one word: Warby Parker does not need its logo for you to remember that blue. Do you have a color, a feeling, a fingerprint that survives without your name attached?

Which is not an argument against AI. Jack is more bullish than most people making that argument. She rebranded her entire company in a single weekend because she could not stop working. She calls the tools a playground and says she cannot put them down. What she wants back is the distinction between the work that deserves a human and the work that does not. “Was I put on this planet to left-align data?”

The business she has built is the answer to that question, and it is stranger than a normal agency. Radiance is an agency plus an intelligence layer. Where most shops pull freelancers in as needed and let the institutional knowledge leave with whoever quits, Radiance is building the memory layer underneath. Her framing of the problem is the sharpest thing in the episode: your logo came from one vendor, your marketing strategy from another, and then Bob decides he wants to become a surfer and walks out with all the mind share. Who was collating, organizing, and storing the brand? Nobody. That gap is the product.

It is a systems instinct she has run four times now. In 1998 she was buying real estate in Williamsburg before anyone thought that was smart. When she realized her freelance art career could not scale around three months of sailing, snowboarding, and surfing a year, she did not just work harder. She went looking for a repeatable model and landed on stationery, because paper is light, does not perish, packs easily, and carries good margins. That became Screech Owl Design in 2007, which reached over 1,800 stores worldwide and, by her account, roughly 2,500 at its pre-COVID peak. The day she launched, Tracey Stewart received one of her products, and Jack ended up creative director of Moomah, the Stewarts' Tribeca creative space. From there came licensing deals with Anthropologie and West Elm.

Then Ollie, the New York co-living company where she led design from 2015 through the Starcity acquisition in December 2020. Same instinct, different asset class: co-living, co-working, hospitality, butler services, roommate matching, all of it about systematizing how people connect. Then Everyrealm, the Andreessen-backed Web3 company, where she was Chief Creative Officer from 2021 to 2024. Then Radiance.

The chapter she is most honest about is the cost. She hit, by her account, $1M ARR by month three. The reward for that was a new question: when are you going to 2X, 3X, 10X? “You get a million, you need a million and one. You get one billion, you need one billion and one.” She calls it the existential problem every CEO deals with, the pull between being present and being permanently future-forward. Her mechanism for surviving it is not a productivity system. It is a 17-person team she describes as kind, inventive, and high-integrity, plus a house mantra: we are either going to glow up or blow up, so let's have fun and love on each other along the way.

There is also a masterclass buried in the middle of this episode on how she actually got the round. She built a side-by-side model: a traditional agency's full-time-employee cost structure against an automated stack running Figma, Canva, HubSpot, and the rest. Put next to each other, revenue per employee stops being an abstraction. She says it may have been the single most pivotal moment with her investors, and she wishes she had built it a year earlier.

She is also clear-eyed about why that chart scares people. It shows exactly how much a company can produce off one person. That is the cannibalization fear in one slide. She does not dodge it.

If there is a moment that explains her, it is on a sailboat on September 11, 2001, with a hand-crank L.L.Bean radio, perfect wind, perfect weather, and a tiny mechanical window into the worst day in New York's modern history. No batteries. No screen. Just the one tool that worked. She uses it now to make a point about auditing your own life: what do you actually love, what technology genuinely improves it, and what is quietly costing you more than it gives back.

Named Frameworks

Digital Darwinism

Her term, and her TEDx thesis.

  • →When everyone can access the same generative tools, access is no longer the advantage.
  • →The selection pressure moves to memorability and distinctiveness.
  • →Survival questions: who evolves, who is sticky, who can change their code, who is still here.
  • →Taste becomes currency, so the richest brand is the most recognizable one.

The Source®, Her Creative OS

The intelligence layer under the agency.

  • →Stores brand assets, messaging, and decisions in one system rather than across vendors.
  • →Templatizes work so a non-creator can operate inside guardrails without depending on the agency for every asset.
  • →Preserves brand memory when a team member leaves.
  • →Sits on top of legacy systems instead of forcing clients to rip and replace.
  • →Turns a six-hour proposal into a seconds-long output, with pricing and scope guardrails already built in.

The Brand Memory Problem

The gap she built the company to close.

  • →Logo from vendor A, strategy from vendor B, campaign from vendor C.
  • →Bob leaves the company and takes the accumulated context with him.
  • →No one owns collating, organizing, or storing the brand's decisions.
  • →The company keeps re-buying knowledge it already paid for.

Agency Plus

Her own two-word category description.

  • →Human creatives stay, because the strategic vision is the part AI does not replace.
  • →AI agents, workflows, and project management sit underneath.
  • →MCP integrations connect the tools clients already run.
  • →The product is the repeatable model, not the individual deliverable.

The FTE Side-By-Side

The investor artifact.

  • →Model a traditional agency's full-time headcount and cost.
  • →Model the same output on an automated platform stack.
  • →Compare capital produced per employee across the two.
  • →Expect the fear response, because the chart makes displacement legible.
  • →Build it early, not late.

The Repeatable Process Test

How she picked stationery over everything else.

  • →Light to ship. Does not perish. Easy to pack. Minimal to produce. Good margins.
  • →The point is not paper. The point is that she chose the business by its logistics before she chose it by its art.

The Technology Audit

Her answer to what actually evolves well.

  • →Start with what you love and care about, not with the tool.
  • →Ask which technology measurably improves that.
  • →Ask what is causing grief and pain, including the notifications you installed on purpose.
  • →Decide, in her framing, whether you want to cut up your fruit or blend it.

Glow Up or Blow Up

The team operating principle.

  • →We have nothing to lose either way.
  • →If we blow up, we were going to blow up anyway.
  • →If we glow up, that is icing.
  • →So build the best work we have ever made and be good to each other on the way there.

Founder Experiment

Run the no-logo recall test. Jack's Warby Parker point is testable in about twenty minutes, and you should do it before your next brand spend. Pull your five closest competitors' sites and your own. Screenshot each homepage. Strip out every logo and wordmark, cropping them out or painting over them. Send the six unlabeled screenshots to ten people who know your category, customers are better than friends, and ask one question only: which of these is us? Log the percentage who get it right. Now do the harder version. Ask them to name your brand's color without looking. Then a feeling. Then one visual thing they could describe to someone else. If recall lands under fifty percent, you do not have a brand problem you can fix with more output. You have a distinctiveness problem, and generating more assets in the same house-of-AI style will make it worse. Pick one element, color, type, motion, or a recurring motif, and commit to it hard enough that it survives without your name attached. Bonus round: build the FTE side-by-side she showed her investors. One column for a traditional headcount model, one for your automated stack. Divide revenue by people in both. That number is your pitch.

Glossary

Creative OS

A single operating system for a brand's creative work, holding assets, rules, and decisions in one place rather than across scattered vendors and files.

The Source®

Radiance's proprietary Creative OS, and the reason the company describes itself as an agency plus an intelligence layer.

Intelligence layer

The software beneath a service business that captures and reuses knowledge, so expertise compounds instead of walking out the door.

Brand memory

The accumulated context of a brand's decisions. Usually stored in people's heads, which is why it disappears during turnover.

Digital Darwinism

Jack's term for competitive selection in a world of universally available AI tools, where distinctiveness rather than access decides who survives.

AI slop

Low-effort, undifferentiated AI-generated output. Jack prefers to frame the problem as taste rather than quality.

Taste

The judgment layer AI does not supply. In her framing, the scarcest asset and the real currency.

Homogeny

The sameness that shows up when everyone prompts the same models with the same defaults.

MCP integrations

Model Context Protocol connections that let AI systems talk to the tools a company already runs, so you can automate without replacing legacy software.

Guardrails

Pre-set constraints inside a workflow, such as pricing rules and scope limits, that make fast automated output safe to send.

MVP

Minimum viable product. The smallest version that tests demand.

Product-market fit

Evidence that a real market wants what you built. Jack's point: AI now lets you find out in weeks instead of after a year of prototyping.

ARR

Annual recurring revenue.

Revenue per employee

Total revenue divided by headcount. The metric at the center of her investor comparison.

FTE

Full-time equivalent. The unit of a traditional agency's cost structure.

Context switching

Moving between finance, roadmap, creation, and client work without losing presence in each. She calls it her single biggest strength.

Licensing

Selling other companies the right to use your designs, versus manufacturing and selling the product yourself.

Wholesale

Selling in bulk to retailers rather than direct to consumers.

Co-living

Shared residential housing with private rooms and communal amenities. The model at Ollie.

Web3

Blockchain-based internet applications. Her Everyrealm chapter.

Camera obscura

The optical device that reshaped painting, and her historical analogy for what AI is doing to design.

Q&A: What Founders Ask After This Episode

How can founders use AI without their brand looking like everyone else's?

Separate the work that scales from the work that differentiates. Jacqueline Schmidt argues you should hand AI the mechanical tasks such as alignment, formatting, resizing, and first-draft asset generation, and keep the strategic vision and taste decisions human. The distinctiveness test is whether a customer can recognize your brand with the logo removed.

What is a Creative OS and why would an agency build one?

A Creative OS is a system that stores a brand's assets, messaging, rules, and decisions in one place and makes them reusable. Radiance built one called The Source® so brand knowledge survives employee turnover, clients can produce on-brand work without depending on the agency for every asset, and proposals that used to take six hours take seconds.

Will AI replace designers?

Schmidt's position is that AI replaces design tasks, not design judgment. Tools now let anyone produce competent-looking output, which has raised the floor and removed access as an advantage. What it has not produced is taste, and that is where she says the remaining value sits.

How do you prove ROI on AI automation to investors?

Build a side-by-side model comparing a traditional full-time-employee cost structure against an automated platform stack, then compare revenue produced per employee across both. Schmidt says this comparison was the most important moment in her investor conversations and she wishes she had built it a year earlier.

What happens to a company's brand knowledge when employees leave?

It usually leaves with them. Schmidt calls this the brand memory problem: the logo came from one vendor, the strategy from another, and the person holding the context resigns. Without a system that collates and stores those decisions, the company re-buys knowledge it already owns.

How do founders handle the pressure after hitting a revenue milestone?

Schmidt hit $1M ARR by month three and says the immediate next question was when she would 10X. Her approach is to build conviction and self-regulation tools before you need them, surround yourself with people who can mirror perspective back, and lean on humor so the growth pressure does not become identity.

Can an artist run a real business?

Schmidt's mother pushed her to art school over business school on the theory that you can start a business at any age but you cannot get creative formation back. She started a company the month she graduated and has now built four. Her framing is that money is an exchange mechanism, not a moral position.

What should founders automate first?

Anything you would not describe as the reason you started the company. Schmidt's test is whether the task requires your judgment. Alignment, formatting, proposal assembly, and repetitive production go first. Client understanding, brand strategy, and taste stay.

URLs Mentioned in the Episode