Skip to main content

Aggregator

Paige Paul’s world of Dairy Boy is collabing with Hanes and coming to Target

1 week 2 days ago

Before the lifestyle brand Dairy Boy had a logo or a single follower, Paige Paul—formerly Paige Lorenze—was a New School student crafting a mood board made up of vlogs and pictures.

In 2021, that vision became the Dairy Boy brand—a name inspired by Paul’s love of dairy products like whole milk, which, in the age of oat milk lattes, seemed radical.

Paul, who now has an audience of more than 2.3 million across platforms, filled the brand with products that showcase pieces of her life. Her upbringing on the East Coast and the changing seasons translate into white eyelet cotton dresses in the summer and colorful fleece jackets in the winter. Then there’s the Instagram-ready pajama sets, and a sporty drop just in time for a US Open run by her husband, the American tennis player Tommy Paul. (He made it as far as the Round of 16 this year.)

“Before I had a brand, I feel like I was building a world,” Paige Paul tells Fast Company.

Now, that world of patterns and Americana lore is growing past the sold-out drops and pop-ups with hour-long lines. Dairy Boy is partnering on a limited-edition collaboration with Hanes that will sell online and at more than 1,200 Target stores nationwide beginning November 14.

[Photo: Courtesy Dairy Boy]

Dairy Boy’s first major collaboration since it launched five years ago pairs Paul’s Americana dream with a staple of that fantasy: a brand built on comfort and functionality just like hers, but rooted in basics that were lacking in her existing collections.

“Everyone in my life, male or woman, has owned something from Hanes, and I think that is what true heritage legacy brand means in America,” Paul says. “Dairy Boy’s collections are so often vibrant and specific, and I felt like the basic side of our business is something that deserved doing it with the best in class.”

Working alongside Hanes and Target allowed Paul to keep pricing affordable. That’s a priority she traced to her days at the New School, the lower Manhattan parent of the Parsons School of Design and other colleges, when she was looking for a price point she could afford while in college.

[Photo: Courtesy Dairy Boy]

The collection keeps the cuts, fabrics, and silhouettes of original Hanes styles, but reimages them through the lens of Dairy Boy. For instance, the collection pairs a classic white tank with men’s boxers, the first time Hanes has sold them specifically for women.

“In my first meetings with Hanes I spoke about how I always stole my husband’s boxers and wore them out to coffee or to bed, and I begged for them to allow me to use their men’s boxers, and they did,” Paul says.

The collection features 45 products, including fluffy bathrobes, waffle pants, long-sleeve sets, a triangle bralette and boy short two-packs, cami sleep dresses, and the husband-inspired boxers. On theme with the drop, small keychains in the shape of a cow wearing boxers will also be available.

[Photo: Courtesy Dairy Boy]

While the silhouettes are classic Hanes, the patterns feature some of Dairy Boy’s most recognizable motifs, including lily of the valley, a brown bear print, camo, plaid, blue and brown pinstripe, and tartan.

The collaboration has been in the works for almost two years, as a silent project that grew alongside the brand’s own successes. Dairy Boy tripled its quarterly orders year over year, surpassing total orders from 2025 in the first nine months of 2026. And those who order seem to come back. In the third quarter of 2026, Dairy Boy reported a year-over-year increase of 312% in repeat costumer volume.

The company has grown, Paul says, thanks to its operations that resemble that of a tech company—integrating technology like AI into their back-end systems and keeping the company team lean. The startup-like structure is paired with Paul’s creative direction, which she says is driven by her gut instinct.

“I am really driving the car here, and if there is no data supporting something, but I really believe in a style, we will order it,” Paul says. “There’s no outside investment, so it’s really easy for us to make decisions quickly.”

Despite being the smaller name in the collaboration, Paul sees the Dairy Boy collection as testament to synergy, pairing the brand’s cultural pull with the infrastructure of legacy companies.

“Target has thousands of stores across America. That is something I don’t have,” she says. “The amount of people that I can get to my website for a drop is something that Target doesn’t have.”

Adobe is fully in its agent era

1 week 2 days ago

I am a big fan of artificial intelligence taking on the miserable tasks of creative work. Renaming and tagging footage, syncing multiple camera angles, reformatting an ad for 50 different aspect ratios? Let the computer do that, thanks. Adobe showed me how its AI agent can do all that, right inside its apps, back in June. But actual creative work? Pass.

Today, at its Max conference, Adobe is announcing something far more radical than anything it showed me just a few months ago. The company is previewing what it calls local MCP integrations with Creative Cloud tools. MCP is the Model Context Protocol, an open standard Anthropic created in late 2024 that acts as a universal adapter, letting any AI agent plug into any software tool.

The integrations would let third-party agents like ChatGPT, Claude, or Gemini work directly inside your local Photoshop, Illustrator, Premiere, and After Effects projects. On your machine. With access to your layers, vector paths, footage bins, and sequences. The projects stay editable by hand, and users will be able to, in Adobe’s words, “step in at any moment.” The apps themselves need to be installed and licensed on your computer, Adobe tells me, because the agent works locally in your project files with the full depth of the desktop tools.

Adobe’s tools already work with third-party AIs—just the other way around. Since December 2025, ChatGPT users have been able to call on Photoshop, Express, and Acrobat to do design work right in OpenAI’s service. And back in April 2026, Adobe launched a connector for Claude to allow it to use more than 50 tools from Photoshop, Illustrator, or Premiere within the Anthropic AI’s cloud-based service. In September, the same ability came to Google.

In all those cases, you drop a file—a PDF, an image, whatever—into a conversation with the AI, ask it to do things to it, and Adobe’s tools will take care of it. The integration allowed users to continue work later in Adobe’s apps.

The new local MCP integrations flip that around, changing who is in control.

[Image: Adobe]

In theory, with Adobe’s own AI agent—now rebranded as Adobe Creative Agent—you know that there’s some degree of control and accountability. Until now, Adobe decided how much its agent could do on its own. With MCP, an outside agent gets the same tools, and Adobe can’t vouch for what it does with them. I don’t know about you, but I feel vertigo. Can we trust these agents with our files? And do we really want to?

It seems that plenty of creators are willing to go further in their use of AI. In the Creators’ Toolkit Report that Adobe released alongside these announcements, a survey of more than 2,000 U.S. creators, 77% say the final creative decision should always remain with the human, even when using agents. They want the help. They want to drive the wheel, too.

But there’s an obvious contradiction between the want and what they get. Some 68% of those using creative AI say they at least sometimes settle for results that don’t fully match their vision. Why? Maybe the answer is that 59% say agents save them time. Or that 36% say agents helped them finish projects that were previously too complex. 

Three altitudes of AI

It’s also happening faster than I could have imagined. Forest Key, Adobe’s VP of Firefly and Agentic AI, told me back in June about what he described as three altitudes of AI. The first is the assistant that strips away drudgery. In an example from that launch we saw an agent in Premiere doing its work: taking a pile of randomly named files, sorting them into bins, renaming and tagging clips using computer vision, syncing footage from three cameras with a separate audio track, and dropping a marker on the timeline wherever an interviewer asked a question.

The second altitude is an AI thought partner, while you keep adding your taste and editorial voice. The third is agents that run full automations while you only check in periodically.

Today, Adobe is also previewing always-on agents, coming first to the Firefly AI Assistant, that take on recurring work in the background—organizing incoming assets, checking quality, preparing content for different channels—while creators set the goal, review the plan, and approve the results. That’s a first step toward altitude three. MCP means agents Adobe didn’t build can climb the same ladder. The local MCP integrations won’t enter beta until early 2027.

[Image: Adobe] Agent as partner

Adobe started its agentic push back in April, when it deployed the Adobe Creative Agent in Firefly, in beta mode. This agent could do the work for you, not just vomit generative AI bitmaps. In June, other Adobe programs also got agentic AI capabilities in beta testing: Premiere, Photoshop, Illustrator, InDesign, and Frame.io. That’s when Adobe’s agent started working on your actual projects from inside its apps. After Effects joined the agentic party in September. 

At Adobe Max, the company is adding new capabilities to the Adobe Creative Agent. In Firefly, the assistant can now remember formats, colors, brand guidelines, and characters to apply across projects. A new Plan Mode lets you review, edit, and approve the agent’s plan before it starts, and a model selector lets you pick which AI model does the work. 

The agent is also coming to Premiere mobile in private beta, where it can turn footage into editable first cuts and make edits with plain-language instructions. In desktop Premiere, it can do really complex things, like apply effects across every clip in a sequence, track masked subjects through a frame, and adjust color and exposure across an entire sequence at once.

Until now, there was no official way for outside AI agents to work inside your local projects like Adobe’s agent can. Whenever it becomes available next year, MCP access to Adobe’s apps for agents like Claude or ChatGPT will truly change things. You won’t bring your work to the third-party agent’s home anymore. The third-party agent will come into your home, get its own bedroom, and half of your fridge.

[Image: Adobe] Lingering questions

Having your own ghost artist is a very tempting yet dangerous proposition. What happens if a third-party agent corrupts a sequence, generates something licensing-toxic inside your project, or quietly burns through generative credits? The accountability chain gets blurry fast. 

Anyone could ask Claude to pull reference imagery from the web, then composite it into your project using Photoshop’s own tools to make a document to open and refine yourself. I doubt Adobe’s own assistant will ever do that. First, many of Adobe’s own customers would hate it. And then, imagine the copyright hell that could ensue.

Asked where liability sits between Adobe, the agent provider, and the creator in that scenario, Adobe only pointed back to its answer that MCP is still a preview. When I asked whether third-party agent actions will carry Content Credentials—the provenance standard the company champions—Adobe also told me that it is “working through how provenance will be supported.” As for credit usage, creators will be able to set parameters and review plans before always-on agents run, but on cost Adobe says it is “exploring ways to provide an upfront estimate.”

Those cost questions aren’t only about outside agents, either. Logically, Adobe’s own assistants already burn credits: The After Effects beta started charging them for its AI Assistant on November 6, 2026.

It just feels to me that, for all the good stuff we get from AI helping with work, there’s still a divide when it comes to actual creative work, one that is full of contradictions at many levels. For now, the open questions seem to outnumber the guardrails.

The strategy question

Then there is the matter about what Adobe gains from opening its walled garden to the AI startup hordes. Why give up the last workflow chokepoint it fully owned? 

Maybe they don’t have any options here. Perhaps this is all about the AI agent craze and MCP becoming the standard. Fighting this wave may be seen as worse—especially by greedy stockholders—than riding it. 

What is clear is that the signs of a company-wide pivot to agents are everywhere. A single Adobe Creative Agent now powers the Firefly assistant, the Creative Cloud assistants, and a new private beta in Premiere mobile, while another service called CX Enterprise Coworker in GenStudio orchestrates agents across brands’ content supply chains, according to the company.

“Many creative professionals already use third-party agents that connect to MCP-enabled apps for parts of their production workflow, and they’ve told us they want those agents to work with the tools they trust to finish the job,” Adobe told me in an email. Translation: The customers were already halfway out the door. Let’s keep them in before they find an alternative, integrate it into their workflow, and start canceling subscriptions.

If MCP works as billed, the change will be profound. Your creative software would become a workshop where you hire whichever AI subcontractor you trust—while Adobe repositions itself not as the brain, but as the bench where everyone works. That is either a masterclass in platform strategy or the beginning of Creative Cloud as someone else’s commodity front-end.

Whatever their business strategy is, this new world feels built on shaky grounds. We are climbing the ladder that Key, Adobe’s agentic AI VP, drew for me in June a lot faster than we can build safety checkpoints and best practices rules. As for me, I’m still happy to let a robot rename my footage. But hand Claude the keys to my Premiere timeline? Not yet.