Bay Area lettering artist Jessica Hische designed the M icon and logotype for Meta's AI assistant Muse.
She drew the icon in a two-day sprint with 32 versions, spent two more weeks on the logotype, and was paid about $12,500.
After she shared the work, many designers on Threads accused her of selling out to AI and Meta.
Hische says she used no AI at all and took the job to keep her two Oakland shops and their staff paid.
Quick read · 2 min
Jessica Hische, a lettering artist in the Bay Area, designed the M logo for Muse, Meta's new AI assistant. She drew it by hand in two days, made 32 versions, then spent two weeks on the matching wordmark. Her fee for the sprint was about $12,500.
When she shared the project in her newsletter, many designers on Threads accused her of selling out to AI and to Meta. Hische says she used no AI, and that she took the job because her client work had slowed and she needed to pay staff at her two Oakland shops.
Meta wanted Muse to feel playful, not cold.
Hische says the backlash left her in tears for days.
She still thinks companies hiring artists over AI is a good thing.
The fight shows how divided creative workers are as AI companies come looking for human talent.
Meta’s new AI assistant has a logo drawn entirely by a human, and the human paid for it online. Jessica Hische, a Bay Area lettering artist, designed the M icon for Muse in a two-day sprint, earned about $12,500, and then watched fellow designers on Threads turn on her for working with Meta on an AI product.
The Muse logo design backlash is a small story with a big tension inside it. Many creative workers see generative AI as a threat to their jobs, and some now treat any work for an AI company as betrayal. Hische’s view is the opposite: a giant tech firm paying an independent artist instead of letting AI generate its branding is exactly what artists should want. Who’s right says a lot about where creative work is heading.
01
How Jessica Hische designed the Muse logo
Hische drew the Muse M in two days and produced 32 versions before the final one. The logotype, the styled word that goes with the icon, took another two weeks.
Meta’s own product designers and engineers had already sketched an M internally but knew it wasn’t right. Their version used a single stroke of the same width all the way around, which looks flat and mechanical. Hische’s specialty is subtle refreshes; she’s redesigned logos for Squier by Fender, Jeni’s Ice Cream and Southern Living magazine. She compares her approach to sending a logo to a spa for a weekend rather than for plastic surgery.
For Muse, that meant adjusting the weight of the strokes where they meet, adding and removing tiny amounts of thickness so the letter feels hand-drawn and warm. Meta’s brief was that Muse should feel creative and playful, not cold and engineered. The team also wanted it to sit comfortably beside Meta’s infinity-loop logo, as if both came from the same family.
One detail got extra attention. Meta’s chief AI officer, Alexandr Wang, joked publicly that he was glad the logo doesn’t look like a butthole, unlike some competitors’ icons. Hische suspects that worry was on the team’s minds, but says her conversations focused on matching the Meta brand.
02
Why the Muse logo turned into a pile-on
Hische announced the project in her newsletter last month with a cheerful headline asking readers not to be mad at her. She acknowledged that people have strong feelings about AI and Meta. The post spread quickly on Threads, where dozens of designers criticized her.
She tried to explain. Her client work had slowed for a year or two, and it funds two brick-and-mortar art and stationery shops in Oakland that barely break even; she says the two made around $10,000 in profit last year. She has four people on payroll, including herself and her mother, and had invested $100,000 in Studioworks, an admin platform for creatives, and $30,000 in Drawling, an art shop for kids.
None of it calmed things down. One critic called the job a matter of ethics, not money, and others speculated about her wealth. A comment about credit card debt for holiday shop inventory was twisted into a claim that she was rich and careless. Hische says she spent two days in bed crying and felt deeply misunderstood.
03
The AI argument underneath it all
Hische expected some pushback, but not this much. She points out that she used no AI at all in the work, and that Meta hired a person instead of generating a logo. In her eyes, that’s a win for artists.
Her critics see it differently. For them, Muse itself is the problem. It’s Meta’s AI agent, software that can do tasks for you such as ordering groceries, calling customer service, sending emails, booking appointments and planning travel. Lending it a friendly face helps sell a technology many artists believe was built on their work without permission.
Muse is also a big, visible product. Meta has unveiled Muse Charm, a keychain-size gadget for talking to the agent without a phone, and plans to bring Muse to its smart glasses, so the logo Hische drew will show up far beyond an app icon.
Hische doesn’t fully disagree. She says the fear about AI’s effect on commercial art is real, and that the anti-AI argument about training on artists’ work without permission has truth to it. She uses AI only for admin and coding help, never to make art, and draws the line at AI replacing human creativity.
04
What this means for creative workers
If you work in design, illustration or writing, this story is a preview of a choice many freelancers will face. AI companies have big budgets and need branding, marketing and interfaces. Taking that money can keep a small business alive, but it may cost you standing in your community.
There are no easy rules yet. Some of Hische’s critics said she should do the work if she needed to but keep quiet about it. She chose to share it because she likes talking about her work and has always been open that her husband worked at Meta for years. For freelancers, the lesson may be practical: decide in advance how you’ll explain a controversial client, because the internet won’t wait for context.
Google's new world model builds explorable scenes with sound in real time. Here is who gets it first, what the Ultra plan costs and what the waitlist means.
OpenAI says its internal models disproved a widely trusted computer science assumption and cracked open problems in math and physics. Experts now have to check the work.
5 min read
We use cookies to understand how readers use Talk With Tech, so we can make it better. Is that OK? Privacy policy
The Daily Brief
Today's biggest tech stories, in 5 minutes
Every morning, the news that matters from AI, phones, apps and the people shaping tech. Explained in plain English. Free.
One email a day. No spam, unsubscribe anytime. Privacy policy