Lately, I’ve been asking myself what it really means to be “AI-first.” Sure, it's a catchy tagline, and the current most overused fundraising buzzword, but it also represents a deep, significant shift in how we work. It feels like so much is happening all at once and reshaping every part of how we build and operate companies.
Figma’s new tools are closing the gap between idea and execution. Startups like Loveable and v0 make it easy to spin up polished frontends from a prompt. Platforms like Replit empower solo builders to launch and scale projects that once needed full-stack teams. Meanwhile, layoffs at major tech companies are being justified under the banner of “efficiency gains” from AI.
Together, these developments tell a deeper story: the collapse of traditional roles.
Who Does What?
The lines between designer, developer, and product manager are dissolving. The assumption that more impact requires more people no longer holds. We’re moving toward smaller, faster teams where AI fills gaps that used to demand entire functions. A team of ten can now ship like a team of fifty. Not because they work harder, but because they work differently.
I’ve seen this firsthand at PermitFlow. In the past few months, I’ve found myself jumping between design, code, and shaping features end-to-end. Thanks to tools like Cursor and Claude (and a new laptop with enough RAM to not shit the bed while running Docker), I've shipped more production code in the last month than I have in the last two years. Even when I'm not submitting PRs, it's still useful to use Claude to "implement" features beyond my engineering capabilities as more of a "live" prototype. I use v0 and Loveable to help me prototype new interaction ideas. ChatGPT helps reformat my "stream of thought" research into our standardized PRD format. The line between design and product management keeps blurring. But, my job title (or my salary) hasn't changed, I'm just doing...more.
This shift isn’t just about speed or efficiency. It’s about how we define roles and where value lives. If design now includes development, a PM can start from a prompt, and engineering can "design" with AI, what does that mean for the "shape" of a company? PMs and designers won’t disappear, but their value won’t be as easily captured in the artifacts they produce anymore.
The Human Aspect
As we move faster and leaner, if feels like we rarely talk about the human impact, even for the folks that aren't laid off. Maybe AI can compress a week of work into a day, but it also introduces new ambiguity. It blurs the line between what the tool does and what you do. Are we valued for what we make, or for how fast we can do it, or how many jobs we can eliminate with AI-based processes? This question sits at the center of how we define ourselves at work. It challenges our confidence, our sense of accomplishment, and for designers, our creative identity.
When roles collapse and structures flatten, people lose more than titles. They lose clear ladders to climb. They lose the comfort of knowing where their contributions begin and end. They lose the rituals that once gave structure to their work...things like reviews, hand-offs, design critiques, stand ups. Perhaps AI enables teams to move at breakneck speed but they often lack a shared language and internal structure to process these shifts and the effect that it has on their co-workers.
For many, this ambiguity will be exciting. But for others, it feels like an existential threat. A designer who once took pride in pixel-perfect visuals now finds AI suggesting entire design directions that frankly...are pretty good. A PM who shaped strategy might watch an AI digest user feedback, come up with feature ideas, prototype, and test flows in minutes. Employees in junior positions now live in fear of getting "AI'ed" completely out of company workflows.
I started my career as a UX designer at a digital agency in San Diego. Some of my work was focused on usability and IA work that would inform full redesigns. These things took time. I had to write stuff down, sketch out information architectures on paper (or in Balsamiq), and spend a lot of time thinking about things. I think that foundation helped me immensely, as now, years later, I can think "fast" about complex design problems.
Now, we can plug a bunch of pages into ChatGPT and ask it spit out an evaluation in 30 seconds. Not necessarily good or complete, but passable. It won't make many of the same conclusions that an experienced designer can jump to, and it can't really capture style or taste. But those things are *learned*. Most companies aren't going to use that time saved to shift priorities and management goals towards new ways coach that same junior designer to a higher level. Instead someone is getting laid off, and someone else gets another task added to their responsibilities because it's now "fast".
A New Shape of Work
It feels obvious that that companies will be smaller. Roles will be harder to define, and "Super ICs" will become the norm. In the coming years, high performing teams may be five to seven people rather than dozens. Hiring will likely focus less on fixed titles and more on hybrid capabilities; prioritizing people who can orchestrate across design, development, and storytelling.
Paradoxically, this shift might lead to more companies, not fewer (Jevons Paradox). As it gets easier to start something, with fewer resouces, more people will. The level of effort to build something meaningful has never been lower.
Yes, AI can collapse timelines and break down silos, but it can’t teach your team how to care (about the product, the customers, or each other). It can’t create belonging, instill a sense of mission, or provide confidence in company strategy. In the end, the companies that succeed won’t just be the most efficient, they’ll be the ones find ways to keep people at the center of the work, even as everything else changes. The human part of building a company won't disappear. If anything, it will become more essential.