Beyond the hype: what “you can just do things” really means for creators
“You can just do things” has become the unofficial motto of the generative AI boom. It started as an offhand meme, but it captures a genuine shift: the barrier between having an idea and shipping something that works has never been lower. Novelists can spin up assistants to refine prose, designers can generate mock copy and prototypes in minutes, and non-programmers can assemble working software from a prompt.
That accessibility is exactly what makes the current moment feel different from previous tech waves. Michael Mignano, a partner at Lightspeed Venture Partners, noted that recommendations for new AI products now come from friends and colleagues outside tech, and arrive on a daily basis. The tools are also unusually inexpensive: premium AI services generally cost around $20 per month, which puts them within reach of a much wider audience. Ben Blumenrose, co-director of Designer Fund, argues this democratization means more people can contribute to product creation than ever before, broadening both the founder pool and the marketplace of ideas that feeds it.
In practice, individual creators are using these tools as collaborators rather than replacements. Peter Yang, a product manager at Roblox, uses Claude to pressure-test his thinking before big meetings, while Jenny Wen, a designer at Anthropic, turns to it for concept clarification and quick prototype iterations.
When “just doing it” goes viral
The rapidity with which ideas spread has also changed. When developer Pieter Levels showed off a flight simulator built in three hours, clones appeared within weeks in every conceivable variation. Vincent van der Meulen, a software engineer at Figma, pointed out that what once would have taken a month to inspire imitators now only takes two weeks.
An even clearer example emerged with WikiTok. After two developers on X discussed an idea for an infinitely scrolling Wikipedia page, app developer Isaac Gemal ran with the concept and had a working MVP live within two hours. That speed catches people’s attention, and it’s amplified by celebrity-fueled trends like the Studio Ghibli–style image craze that recently swept social media. But the viral nature of creation cuts both ways: when anyone executes on the same fleeting prompt at the same speed, novelty disintegrates into sameness almost instantly.
There is also a mismatch between the thrill of shipping fast and the quality of what gets shipped. As Benji Taylor, chief product officer at Avara, put it: you can do too many things and end up completing nothing at all. Robyn Park, head of platform at Designer Fund, surveyed more than 400 designers about their AI habits and found that the overwhelming majority were self-taught, with no formal training. The result is a widening gap between those who experiment actively and race ahead, and those who wait for standards to coalesce and risk being left behind. Microsoft’s Future of Work Report 2024 adds a cautionary note: AI may improve performance on some tasks while inflating users’ confidence, leading them to overestimate what they are truly capable of on their own.
That distinction matters. “Here’s the uncomfortable truth: Speed isn’t the same as quality,” says Sara Vienna, chief design officer at Metalab. The best work emerges when people know what to reject, refine, and reimagine. Nikolas Klein, a product manager at Figma, agrees that you still need a coherent vision to prompt toward, which requires trying things out, iterating, failing, and starting again.
Lessons from the no-code movement
This cycle has precedent. The no-code movement of the 2010s was built atop visual programming languages and fourth-generation languages from the 1970s and ’80s, and its pitch sounds eerily familiar: creative people would no longer need years of engineering expertise to turn ideas into products. In practice, many of those platforms demanded their own steep learning curves. David Kossnick, senior director of product AI at Figma, sees direct parallels in today’s codegen tools. They have certainly opened new categories of makers, but “no-code” never meant “low effort,” and generating software from plain-language prompts does not automatically eliminate the hard thinking that software actually requires.
What this shift means for work — and job titles
There is no putting that genie back in the bottle. As AI tools flow into the workplace, they are blurring the traditional lines that separated design, development, product management, and writing.
The adaptive challenge is no longer purely about output. Ben Blumenrose notes that people are unsure how to rework their entire tool stacks around this new approach to building. Basic questions remain unresolved, including where AI-built prototypes should be stored, how teams give feedback on them, and what role non-technical designers now play in the loop. Andrew Chen, a general partner at a16z, has described the successful operator of 2025 as a triple threat: content creator, vibe coder, and holder of an actual job.
The skills that differentiate individuals in this environment are also shifting. Vienna argues that taste, not proficiency, is now the key asset: the creators who thrive are those who know not just how to use these tools but when to use them. Kossnick points out that we are still in the earliest days of AI interface design, and compares today’s prompt boxes to the command line before graphical user interfaces arrived. The experimentation happening now, he says, spans both core model capabilities and the interfaces that will eventually determine whether tools such as these become broadly accessible to more people — rather than just early adopters or experienced builders already accustomed to navigating new software.
Experimentation is the point
There’s no question that AI has supercharged the industry’s already strong bias toward action. Teams are shipping faster, iterating more, and exploring ideas that would have been impractical to prototype just a few years ago. “Just doing things” has become a genuine creative strategy, not just a slogan.
But speed without direction isn’t inherently valuable. The tools we’re using are expanding what’s possible, yet they don’t answer the harder question of whether we’re building the right things. The cultural momentum around rapid experimentation needs a counterweight: intentionality about what deserves the effort in the first place.
That’s where the historical record offers some reassurance. Time and again, the most significant breakthroughs have come not in the first wave of a new technology, but later—after the tools mature and people figure out workflows that genuinely leverage them. Early adoption is useful, but it’s rarely where the real value lands.
The implication is straightforward: the current pace of experimentation is fine—even desirable—but it’s not an end in itself. Steering these tools means keeping the bigger picture in view while the details get worked out. If past cycles are any indication, the most meaningful innovations from this moment are still ahead of us, and the teams that balance speed with judgment will be the ones to find them.
Andrew Hogan leads Insights at Figma. His research focuses on the digital product and design industry and the ways the most successful teams work. Previously, Andrew spent seven years at Forrester, a leading research firm, analyzing the intersection of design and tech.
