Dylan Field: Is AI Replacing Designers?, Design is the New Code & Democratizing Creativity
In a Nutshell
Dylan Field argues that as AI commoditizes code, design becomes the key differentiator—enabling brands to stand out through bold creativity, emotional resonance, and unique points of view rather than technical execution. While AI democratizes creation by lowering the floor for production, humans raise the ceiling by operating outside training data distributions and solving complex, multi-level problems that combine user needs with marketing, brand, and cultural systems. The Adobe acquisition's failure and AI's rapid model cycles demonstrate that founder-led companies focused on user experience and rapid iteration will outperform those relying on AI alone, as transparency, intention, and the ability to craft AI outputs into something genuinely new remain essential competitive advantages.
These notes were generated by AI and may contain inaccuracies.
Creative Planning helps manage complex wealth involving investments, tax strategy, and estate planning through coordinated teams of investment managers, CPAs, and affiliated estate attorneys.
Dylan Field argues that design is the new code. As coding models improve, the nature of programming changes and the value rises in differentiation at the top - this includes design, brand, marketing, being bold, and eliciting emotional reactions.
Figma was founded in August 2012. The goal is to enable the best teams worldwide to design, create, and build the world's best software. Figma is used for marketing, advertising, slides, whiteboarding, diagramming, and any visual work.
Design is Figma's core. The definition includes creativity meeting problem solving. Design spans interior, industrial, product, brand design and more.
Adobe offered to buy Figma for $20 billion. The deal was signed but blocked by antitrust authorities from the United States, European Union, and CMA. The regulatory review lasted 16 months.
One host used Photoshop since 1990, doing graphic design pre-layers as a kid on the original 1984 Macintosh. They used early 3D tools like Bryce 3D and Kai's Power Tools, which expanded creative potential.
The deal was blocked by multiple governments. The company kept building despite uncertainty. After 16 months of regulatory review, an all-hands was called on the first day of winter break. The mood was relief from ambiguity rather than disappointment about the deal not closing.
Roomba's acquisition was blocked, leading to bankruptcy and the market shifting to China, demonstrating how antitrust decisions can harm American interests.
Markets are now more expansive. Deals that appear large in historical market definitions may represent 1-2% of the actual future market, particularly with AI expanding opportunities.
The company offered severance packages to employees who joined expecting an Adobe acquisition and preferred the startup pace. A small percentage left, some making career pivots. After 6 months, former employees could reapply.
Figma partnered with Anthropic through an MCP connector. When Anthropic announced Claude Design separately, it caused market speculation that Figma was finished, though business continued growing at 48% year-over-year.
Not all SaaS companies are equal. Founder-led companies with embedded user experience and ability to use AI to accelerate business perform better. Cursor exemplifies this by obsessing over users and seeing around corners despite competition from frontier labs.
AI lowers the floor for digital production including code, motion, video, and design, democratizing creation. However, AI does not raise the ceiling - humans raise the ceiling through bold creativity and doing what hasn't been done before.
With AI enabling exponential software creation, usage remains flat while app numbers grow. Differentiation comes through design, brand, marketing, and point of view rather than code, which anyone can now create.
The point of view is often missed. It's the human part that involves being uncomfortable saying something outside consensus. Comedians demonstrate this by speaking truth that others cannot hear.
Social media diffuses unique points of view instantly. Messages must be evergreen and outside the overton window to break through sustained attention.
Creating something new requires doing what hasn't been done before, operating outside training data distribution. This uniqueness is rewarded by markets and attention.
Figma requires AI-generated writing to be labeled "Written with AI" at the top, creating trust through transparency.
AI helps overcome blank page problems. Users can start with AI output, then react and modify based on their vision. Having a vision of how something should look or feel allows meaningful iteration beyond AI output.
If you have your own vision and you're trusting AI to implement it, you're doing it through the lens of AI, and AI has its own distribution and model fit. You're not going to get exactly what you want. The goal at Figma is to make code a material on the canvas so you can start by building, build to think, and also work with your hands when you want to, going back and forth rapidly between modes.
Early Photoshop users reacted strongly when layers were introduced, then again when filters and Kai's Power Tools emerged, with concerns about destroying the art and creativity of photography. However, these new tools unleashed a new era of creativity as artists emerged who could use filters, layering tools, and capabilities to reimagine what's possible on a canvas. This represented a new toolkit where old creatives didn't map perfectly to the new creative system, but there was an explosion in creativity and human potential that pushed boundaries in all sorts of ways.
Market dynamics require doing something new to get attention, engagement, and care, which means being daring and trying bold things. The resistance to change is real, as evidenced by the publicly available mailing list reactions when layers were introduced, where people called it the worst thing ever. Despite initial resistance, layers became the future. While photographers still do incredible work, the bigger disruption was the camera phone, which made dedicated cameras obsolete because the best camera is the one you have with you, transforming photography through Instagram and changing how people treat the medium.
One view holds that human creativity is connecting dots in the unconscious from sensory inputs and experiences, then extrapolating a forward view of things not seen before. Humans are always seeking novelty with different reward and loss functions than AI. For example, splashing water would make someone avoid doing that again, whereas AI requires multiple corrections through reinforcement learning. The brain is fundamentally different from current AI systems, and neuroscience remains largely unknown. Design differs from pure creativity as it involves problem-solving combined with creative elements.
Design requires solving user needs in a creative way that stands out and connects with people, while also connecting to wider systems including marketing, brand, culture, and business problems. Holding all these elements at different levels of abstraction while delivering a screen or flow creates a combinatorial explosion from a search-based standpoint.
AI enables reinventing existing art forms and media through more dynamic generation where creators step up a level, set parameters, but don't necessarily render the final product. Like video games, the final product gets rendered on the fly with the user present, while creative teams build objects, storylines, and narrative flow. This approach will likely spread across many different media types.
Figma Design includes generative tools and shaders for dynamic work, plus a product called Weave that provides a node-based system for connecting model outputs to other inputs. The concept is that generating with AI isn't bad, but you need ways to craft and mold that output. Users create amazing video and photo work through Weave, achieving incredible outcomes when they have intention and care to do the work.
When new models drop, there's an AI psychosis moment lasting a few days aided by algorithm feeds where the best results rise to the top. People initially think the model is everything, then adapt within days as they recognize the model's main tropes and the novelty fades. Models show strong progress in formal domains like math and cybersecurity with stacking IQ points, but creativity and design show incremental progress, possibly due to their non-verifiable nature.
Notable AI experiments include a designer artist who created a webpage as a letter to Figma that gradually degraded each time viewers moved away and returned, serving as a comment on attention. This technique was considered neat and unique. Earlier examples include NFT art experiments from around 2019 using AI, and Holly Herndon's work making her voice available in an almost open-source way for others to use, which was ahead of its time and explored the intersection of what is individual versus collective, unique and ownable versus not.
Negative AI commentary suggests it takes away human agency and caps human potential. However, the choice remains to be a consumer or creator, to take the easy path or climb the mountain. Many people are letting AI think for them rather than using it to further understanding, learn things that would have taken too much time, or expand human potential. The goal is using AI to create good and be a force for positive change rather than joining the hive mind.
Everyone can be a maker and do bold things through willpower and curiosity. Being curious and willing to do the work enables incredible things. The future will likely feature a vastly different setup where unimaginable things become possible and current daily activities are no longer in the human realm, though the exact form remains unclear.
The OpenAI math drop released over 700 papers on famous problems, representing the greatest day of discovery in human history. This created mixed emotions of brilliance and excitement alongside sadness about simultaneously losing and gaining something. The frontiers appear brighter, wider, and more accessible than ever. Math holds special mystery as an unsolved problem, and while progress is exciting, mourning the loss of some mystery isn't wrong.
Keep All-In Podcast in your library
Save the videos and channels worth coming back to, and find them again in one place.
