Good morning. What do designer leather handbags have to do with AI?

That’s the question Coach and Kate Spade owner Tapestry tried to answer this week when it hosted its second annual AI summit: a 3-day event aimed at drumming up internal excitement for its many AI use cases.

Employees from across the company convened at its Hudson Yards headquarters in New York for presentations, talks and a hackathon-style competition at an event Tapestry Chief Information and Digital Officer Yang Lu said was helping shape a culture of embracing AI.

“The biggest thing we have to continue to evolve is culture. AI is not a technology thing. It’s a culture thing. It’s about people,” Lu told me when I stopped by the event on Tuesday.

And I can attest that the atmosphere was lively, a sign that this company is moving into the AI age. Here’s how:

Understanding Gen Z

Tapestry’s consumer insights team is constantly gathering data and information about their customer base. They send out surveys, devour reports and will even go to customers’ houses and look through their closet (Seriously).

But there was a bottleneck in terms of distributing that information to the designers, marketers and sales teams that the data was meant to help. So the consumer insights team used their vast pool of data to create a series of customer profiles including Gen Z shoppers from the U.S. and China.

Gen Z is a critical segment for Tapestry. “It’s our core,” Lu told me.

Now employees can have chatbot style conversations with these different personas, asking questions not just about their shopping habits, but also about broader areas like their lifestyle, goals and ambitions. One Gen Z persona told me that she was worried about the job market, which only felt a little ironic.

A New Way to Coach Sales Associates

Last year, Kate Spade overhauled its approach to what it now calls the “selling ceremony,” an experience between in-store shoppers and the sales associates who can nudge them into making a purchase.

After training store associates on new best practices, the company turned to AI. 

Associates can now have a live voice conversation with an AI-generated persona of a shopper, practicing how and what to say to close the deal. The AI app will then generate feedback based on how they performed and give them a score out of 10. 

Victoria Santoriello VP Customer Experience & Go to Market Strategy at Kate Spade, told me it’s too early to tell whether associates who took more trainings or scored higher actually closed more sales, but the anecdotal feedback is promising. 

She noted that associates used to rehearse customer-advising scenarios improv-style with colleagues before the store opened. They much prefer the AI experience, she said.

Designing With AI

As designers tap in to AI-powered tools to conceive and test new products faster, real-life physical prototypes remain a critical element of the process. That’s because a team of on-site craftsmen must work hands-on with them to validate the designs and make sure everything looks and feels up to snuff.

“The AI doesn’t know that certain piping can’t be engineered in the way that the AI generated it. The craftspeople know that, so they can know that on a design, and can provide a suggestion to move forward,” Tomer Emmar, senior manager, Digital Product Creation, Tapestry told me. 

Emmar said the company now has half a dozen 3-D printers on site that print everything from small accessories handbags to sneakers. The timeline for 3-D printing a shoe is 1-3 days, which he said is a 75% reduction compared with ordering a prototype from an overseas factory. 

“I would say that in the beginning, it was a little bit difficult to help people understand the value of 3-D printing and it took some time,” Emmer said, “but also over the years, the culture of innovation changed and people became more open.”

Meta Platforms on Wednesday announced a new coding agent, called Muse Code, with a discounted tier priced near Chinese competitors like DeepSeek in exchange for user feedback, the Journal reports. The launch comes as Meta faces investor pressure to justify its heavy AI spending after shares dropped 10% last week following a lackluster earnings call. Anthropic’s Claude Code took off in December and OpenAI unveiled its own coding agent, called Codex, in May.

Google Chief Scientist Jeff Dean is leaving to co-found an independent company, Discovery Loop, along with Sanjay Ghemawat—who worked alongside Dean to build Google’s distributed systems in the early 2000s—as well as Quoc Le, a founding member of Google Brain, and Oriol Vinyals, a lead contributor to AlphaFold. The startup aims to accelerate scientific and engineering progress by automating complex, multistep research tasks, the WSJ reports.

Also Wednesday, Google CEO Sundar Pichai announced that Demis Hassabis is stepping down as chief executive of Google DeepMind to become the unit’s chairman and Alphabet’s chief scientist. Koray Kavukcuoglu, currently DeepMind’s chief technology officer, will take over as senior vice president of the unit, reporting directly to Pichai.

Chinese AI models might be winning the AI race in Africa with developers there favoring the cheap and available models over more powerful (and more expensive) U.S. versions, the New York Times reports. Developers tell the NYTimes that Chinese companies are sweetening deals by providing personal incentives to switch, drawing on the years of on-the-ground experience China’s tech and telecom providers have built up on the continent.

Everyone now is an AI chip developer. The Business Insider reports that Anthropic is on the hunt for engineers to design custom chips for its Claude AI models. This comes after rival OpenAI and Broadcom in June announced Jalapeño, a custom chip for model inference. Meta is undertaking its own chip program, while Amazon, Google and Microsoft all house their own custom AI chip operations.