Why cultivating agency matters more than cultivating skills in the AI era | Max Schoening (Head of Product, Notion)

Notion designers and PMs started coding to prototype AI interfaces

Notion's product teams began coding out of necessity when traditional design tools like Figma proved inadequate for prototyping interactive AI chat interfaces. Max Schoening explained that static designs were like "drawing dead fish" because they failed to convey the dynamic feel and real-time interaction required when working with AI.

To address this, two designers, including Schoening, built a simple, LLM-friendly coding playground. This environment made it easy for non-engineers to experiment, requiring minimal interaction beyond the terminal. It allowed teams to rapidly prototype chat UIs and recreate patterns aligned with existing model capabilities.

This initial playground successfully introduced designers and product managers to coding for AI. It served as a low-barrier entry, helping them overcome the fear of the terminal and directly engage with AI prototyping. The experience has since evolved, with these teams now contributing to the production codebase as AI model capabilities become more sophisticated, indicating a shift towards blended roles.

The static image of a chat is basically the dead fish here. you have to feel the, the AI to some degree.

Notion Designers Primarily Prototype in Code and Contribute to Production

Moving designers from Figma manipulation into code has always been beneficial. Max Schoening observes that at Notion, designers now primarily code and prototype directly in code. He recalls that at GitHub, before large language models, designers accounted for 10% of the top contributors to the platform.

One emerging issue is that designers creating prototypes in code are sometimes asked by marketing teams to reverse-engineer those designs back into Figma to create assets for videos. This creates unnecessary busy work.

On the spectrum of pushing to production, designers commonly make small changes, such as styling tweaks. However, Schoening expresses a concern about "vibe coding," noting that while the quantity of software has increased significantly in the past year, its overall quality and reliability have not kept pace.

The true value of designers coding isn't merely about them deploying to production. Instead, it's about enabling them to think and design within the actual medium that will ultimately become the final product, ensuring a more direct and effective design process.

I don't feel like the quality of software has increased all that much in the last twelve months. I think maybe the amount of software has, but it's very, very hard to find software that is reliable.

The Value of Coding for Deeper Design Understanding

Max Schoening challenges the common debate about whether designers and PMs should write production code. He asserts that the primary value of coding for these roles is not about shipping features faster, but about gaining a profound understanding of the underlying technical material they are working with. This perspective shifts the focus from delivery efficiency to design efficacy.

He emphasizes the importance of understanding "agent loops," which he considers fundamental to current software development. Schoening argues that without hands-on experience building these components in code, designers and PMs can only grasp a superficial understanding. This deep engagement with the material is essential for designing effective and robust solutions, especially as software increasingly involves complex, self-contained systems.

For example, Schoening prefers a designer or PM who can deeply understand and design agent loops, even if they can't tweak UI styles, over someone who can only handle surface-level software tasks. Coding, in this context, serves as a tool for interrogation and mastery of the medium, allowing professionals to become experts in their design material rather than merely contributing to the shipping process.

I think it's really important not to forget that the reason why is to become a master of the material, not a sort of cog in the delivery.

Agency is the ultimate differentiator for success in the AI era

In the age of AI, where advanced models increasingly augment traditional skill sets, success for professionals like product managers and designers no longer solely relies on technical ability. Instead, the ultimate differentiator is agency—the belief in one's capacity to influence and change the world around them.

Max Schoening explains that while a 'skill issue' might have been a convenient excuse in the past, AI's ability to assist means that even with skills at one's fingertips, agency becomes paramount. This mindset enables individuals to view their environment as malleable, rather than a fixed system they must simply adapt to.

Those who possess true agency and understand the world can be shaped by their actions are poised to thrive. Conversely, individuals who rigidly adhere to predefined job roles and responsibilities, rather than seeking to innovate and adapt, will face significant challenges in the evolving future of work. Cultivating this sense of agency is therefore essential for navigating the AI era successfully.

I think people who have true agency and they understand that the world around them is malleable will do great.

Cultivating Agency and Its Impact at Notion

Max Schoening observes that Notion employees, or Notinos, demonstrate an exceptionally high level of agency compared to other organizations. This means individuals consistently take initiative and blur traditional role boundaries to drive impact, even within an established company.

Brian Levin exemplifies this by not only blending engineering and design responsibilities but also serving as a top recruiter. He actively seeks to affect change beyond his day-to-day duties, demonstrating a proactive approach to company growth and product development.

Another example is Eric Liu, who evolved from writing strategy documents to building prototypes. After asking if he'd be hired early in a startup, he intentionally developed design and prototyping skills in Figma, shifting from conceptual work to direct product contribution.

These examples highlight a culture where employees feel empowered to contribute significantly and creatively as if they were founders, rather than merely adhering to their job descriptions.

Do you drive Notion like it's stolen?

Merging Roles Risk Losing Specialized Engineering and Design Expertise

As software roles like engineer, product manager, and designer begin to merge, there is a risk of losing specialized expertise. The industry traditionally benefited from clear distinctions in these roles, but with evolving technology and "malleable software," there's a concern about the erosion of clear career paths and design consistency.

Max Schoening illustrates this with a physical product metaphor: while prototypes might be quickly 3D-printed with visible layer lines, true engineering ensures a product can be manufactured with precision for millions of users. In software, this deep engineering often gets overshadowed by a focus on shipping features and token spending, neglecting the meticulous work required to ensure robust functionality for a massive user base.

Similarly, in design, accessible design systems allow for quickly building usable user interfaces. However, this accessibility might lead to a neglect of the specialized craft and "delight" that expert designers bring. The concern is that as roles blend, the unique contributions of specialists at the

edges

I think if we're not careful, we will lose Specialists.

Develop Personal Agency by Consistently Making and Tinkering

Developing personal agency often starts with the act of making and tinkering. This hands-on process allows individuals to realize that many things in the world, from software to furniture, are created by people who are not inherently smarter than them. This realization breaks down barriers and empowers people to believe they can also build and change things.

The guest, Max Schoening, emphasizes that creating and tweaking items, whether it's learning to build a chair, modifying something in your office, or even cooking a meal, places you on a 'treadmill of creation'. This continuous engagement with making is an innately human activity, akin to toolmaking or creating art.

This approach contrasts with the common misconception that agency is about navigating corporate structures or circumventing difficult bosses. Instead, the focus should be on building skills and creating tangible outputs. When people consistently make things and improve, they often naturally attract attention, which further reinforces their sense of agency and capability.

One day you wake up and you realize the world is made up by people no smarter than you.

The Concept of Malleable Software and User Ownership

Malleable software refers to tools designed to prioritize user interests over those of the creating corporation. It empowers individuals to change and customize the software extensively, moving beyond rigid, pre-defined user experiences.

The current software landscape, dominated by apps, is compared to living in a home where rearranging your living room or kitchen is forbidden. Each app's user interface and data ownership are tightly integrated, making even minor modifications impossible. While extreme solutions like running a custom Linux distribution offer malleability, they come with the burden of constantly building and maintaining everything from scratch.

An increasing number of people are building their own tools, representing a form of malleable software. However, for this approach to scale and genuinely empower users, it requires underlying platforms and operating systems that encourage communal development and sharing. This fosters collective ownership and collaboration, rather than isolated individual efforts, to advance the concept effectively.

Imagine you lived in an environment where you do not get to rearrange your living room, and the kitchen has to be exactly set up the way that someone else decided. We would not take that, right? But that is kind of the world that we have in software right now.

Dieter Rams' Philosophy: Design for Usefulness Over Beauty

Max Schoening connects a video of German designer Dieter Rams criticizing impractical chairs to his own design philosophy. Rams dismisses designs that appear beautiful but are not functional, especially those created for display rather than actual use.

Schoening aligns with Rams' belief that design's primary goal should be usefulness, with beauty as a secondary consideration. He contrasts this with "museum pieces" that might look impressive but are uncomfortable or nonsensical in practical application.

This emphasis on utility ties into the concept of "malleable software" and adaptable systems. Just as software can be changed, physical designs should also be tweakable. The ability to modify a design over time is a key indicator of its long-term usefulness and ability to adapt to user needs.

Schoening also notes that the ability to change and tweak a design is a good indicator of its usefulness. This concept applies to things like homes that adapt over time to how people live, rather than being rigidly designed from the start by an architect.

design should be first useful and then beautiful.

Debunking the 'SaaS apocalypse' by focusing on the 'as a service' value

The idea of a 'SaaS apocalypse,' where users forgo tools like Notion to rebuild everything themselves, is largely exaggerated. Most individuals and companies do not want the burden of maintaining complex software stacks. Max Schoening's personal experience attempting to rebuild Notion in a weekend highlights the significant frustration and ongoing effort involved.

The enduring value of SaaS lies in its 'as a service' component, which provides continuous maintenance and specialized expertise. Software, as Brett Taylor suggests, is like a garden that requires constant tending. Users effectively pay for a team of specialists to dedicate their focus to a problem, freeing them from the constant upkeep.

While tools like Notion are evolving to become more general and malleable, especially with the integration of AI, they will still operate as a service. This evolution allows users to build more complex solutions without shouldering the full maintenance responsibility, as demonstrated by Notion AI making the platform more accessible.

Even highly advanced companies like Anthropic, despite their technical prowess, heavily rely on established services such as Slack rather than building their own internal communication tools. This illustrates a broader principle: organizations prefer to invest their time and resources into core competencies, outsourcing complex, non-differentiating services to specialists.

Software is like a garden; you need to tend to it, and what you pay for in the as a service is a bunch of specialists thinking really hard about a problem.

AI makes the initial 10% of projects free, enabling rapid exploration

AI has fundamentally altered the initial stages of product development, making the first 10% of any project essentially "free." This means that teams no longer need to spend extensive time on upfront documentation, such as Product Requirements Documents (PRDs), before seeing a tangible product.

Instead of writing lengthy specifications, developers can now quickly build a "janky version" or even "version zero point eight" of a startup or feature. This dramatically lowers the barrier to entry and exploration, allowing teams to prototype ideas much faster than before.

This shift enables organizations to explore multiple paths concurrently and affordably. Teams can "send off ten agents to explore ten different things" to test hypotheses and gather early feedback. The emphasis moves from theoretical "memos" to concrete "demos."

The goal is to provide stakeholders with something tangible to react to much earlier in the development cycle. This accelerates the feedback loop and allows for quicker iteration, ensuring that product development is more responsive and less reliant on speculative planning.

The first ten percent of every project are now free.

Future Shifts in Product Building: Plaintext, Agents, and Instant Inference

Max Schoening discusses the tension between the enduring reliability of plaintext and markdown for expressing code and the rising prominence of chat-based AI interfaces. He considers whether direct manipulation in tools like Figma will diminish as agents increasingly handle design tasks, noting varied adoption of AI tools among designers at Notion.

The speed of AI inference is a critical factor in future workflows. Currently, slow inference leads to queuing tasks and fragmented attention. However, Schoening posits that nearly instant inference could restore direct manipulation, allowing users to fluidly "mold the clay that is the code" in real-time, potentially fostering a more focused flow state.

Schoening questions the prevailing assumption that users will always demand the most intelligent AI model. For many knowledge work tasks, he suggests that a "good enough" level of intelligence, when combined with optimizations for speed, cost, and local operation, might be more beneficial than pushing for absolute frontier intelligence. He draws an analogy to retina displays, where additional pixel density becomes irrelevant beyond a certain point.

He believes society's limitations are not solely due to a lack of intelligence. Instead, the focus should shift towards developing "exoskeleton" AI tools that augment human capabilities through efficiency and accessibility, rather than a "god in a box" model that exists remotely in a data center.

If the inference becomes instant Do we get back to direct manipulation, right? Do you, you instantly sort of like mold the clay that is the code, right?

Notion allows unlimited AI token spend but anticipates future ROI scrutiny

Notion currently maintains an "unlimited" policy for AI token spend, especially for tasks initiated by humans, aiming to encourage employees to explore and experiment with new AI capabilities. While specific figures are not closely tracked by product leadership, it's acknowledged that individual monthly spend can be substantial, potentially reaching thousands or tens of thousands.

Despite the current open-ended approach, it is predicted that within the next 6-12 months, many companies will start having "uncomfortable conversations" about the actual return on investment (ROI) from their AI usage. The current luxury of not scrutinizing costs is expected to diminish as AI adoption matures.

The focus should not be on token spend as a vanity metric, much like boasting about lines of code written. Instead, the real challenge lies in integrating AI agents into employees' outer loop workflows and creating what's sometimes called a "software factory." Getting people to adopt these new ways of working often requires significant prodding, especially in large organizations like Meta, which has resorted to leaderboards to incentivize adoption.

I do suspect in six to twelve months from now, a lot of companies are going to actually start asking questions around ROI And I think that will be an uncomfortable conversation for, for a lot of folks.

Encouraging Workflow Changes and AI Adoption Across Roles

Non-engineering roles at Notion readily adopt AI because it dramatically increases their capabilities, giving them an immediate "superpower" feeling. The gap between their previous abilities and what AI enables is so vast that it's an intoxicating experience.

In contrast, engineers face a different challenge with AI. The focus for engineers is on rethinking manual interventions in code. Simon Last suggests that any manual intervention during code creation, beyond review, often indicates a problem within the verifiability loop or the overall software factory.

To encourage significant changes in established workflows, organizations need to over-index on shifting people's default behaviors. This is crucial because people naturally gravitate towards familiar routines, even when new, more efficient tools are available.

Every time a human intervenes manually in the code writing process, it should serve as a litmus test, prompting questions about potential inefficiencies or issues in the automated system.

Whoa, I have superpowers now. Look at this amazing thing I've just built.

Max Schoening discusses the increasing use of terminal-based coding for designers and PMs at Notion, moving away from relying solely on GUIs.

Max Schoening observes a trend at Notion where designers and product managers are increasingly adopting developer tools, specifically the terminal, over traditional graphical user interfaces like Figma. He clarifies that this isn't necessarily Figma "trending down" but rather a parallel rise in terminal use.

He actively encourages PMs and designers to use terminal-based tools, even if initially intimidating, such as Cloud Code or Codex. His reasoning is that direct interaction with these tools fosters curiosity and helps them gain a deeper understanding of the underlying computational substrate.

This shift means that many teams are now primarily utilizing developer-centric tools, blurring the lines between roles in their tool stack and making them not much different from what developers use.

I generally encourage them to not use the GUIs. I encourage them to use the tools because I just know that over time they're gonna be curious And one day they wake up and they're like, 'Oh, I understand more of the substrate of what, how, how computers work.'

AI Accelerates Software's Expansion into All Domains

AI's next major transformation will not be in a specific non-engineering role like marketing or HR, but rather in fundamentally changing the cost and accessibility of software development. Max Schoening argues that AI is making coding exponentially cheaper, which is rapidly altering the landscape of software engineering itself; for instance, almost all code could be AI-generated within a year.

The core idea is that as the cost of creating software and encoding business practices in code approaches zero, the volume and application of software will dramatically increase. This isn't about AI becoming significantly better at general non-code writing, which Max finds still tends to be 'AI slop writing', but about its profound improvement in generating functional code.

This reduction in coding cost means that software engineering principles will increasingly permeate all other domains. Teams in departments like HR or marketing will be able to automate tasks and build custom tools without needing to rely on a dedicated engineering team, as the barrier to creating software components drops significantly.

Essentially, AI's primary impact is accelerating the existing trend of 'software eating the world.' This perspective is reinforced by observations that even when AI models show 'progress' in non-coding domains, it often involves applying coding principles, with winning 'agents' typically being coding agents that build what they need on the fly.

If the cost of software and creating software and encoding business practices in code goes to zero, we will just have a lot more of it.

Companies will soon prioritize ROI in AI adoption, favoring smaller models if the gap between frontier and open-weight models doesn't widen significantly.

Max Schoening predicts that within about six months, companies will begin to scrutinize the cost-effectiveness of AI adoption. This shift will involve rigorous ROI calculations, potentially leading businesses to favor smaller, more affordable AI models over larger, frontier ones.

This strategic choice hinges on the performance difference between advanced lab-developed models and open-weight alternatives. If the performance gap remains narrow, businesses will become comfortable running and fine-tuning their own models, similar to current practices at companies like Cursor, Intercom, Notion, and Unit, where 'good enough' AI is sufficient for many tasks.

Max compares the impending AI market dynamics to the 'cloud wars,' where businesses eventually opted for flexibility and avoided vendor lock-in. He believes companies will prioritize choice and cost efficiency, seeking solutions that empower their teams rather than tying them to a single AI provider.

The current competitive landscape, with multiple competent labs (at least three in the US, potentially more with Meta) developing frontier models, is viewed as a fortunate scenario. This competition prevents centralization of power and offers businesses more options, fostering an environment where innovation thrives through choice.

Notion AI's Success Through Context-Rich Connected Workspaces

Notion AI's unexpected success is attributed to its native context-rich environment, which provides AI agents with the necessary information to operate effectively. Unlike systems with data silos, Notion's connected workspace allows agents to access and process information seamlessly.

Max Schoening highlights that agents thrive when they can 'roam around' a unified data repository, making Notion's structure inherently valuable for AI. This environment allows AI to function more like coding agents in a Unix-like operating system, where context is readily available.

The platform embodies the principles of malleable software, meaning it can be adapted and configured by users, which further enhances its utility for AI integration. Notion's long-standing presence as a comprehensive repository for company data made it a perfect fit for leveraging AI, providing a rich source of context for various tasks.

This success demonstrates the critical role of a well-integrated data environment in enabling powerful AI applications, rather than just the AI models themselves. Notion's ability to provide broad contextual understanding allows its AI features to be genuinely useful to a wide range of users.

The fact that Notion basically has all the things about everything in your company is the perfect source of context for using AI and helping you work.

Balancing Shipping Velocity with Software Quality

Max Schoening observes that established companies like GitHub and Notion can develop a "preciousness," leading to internal debates about what to ship next after initial success. He advocates for reducing this hesitation, reminding teams that it's okay to "just do stuff" and be less precious about every release.

An internal metric at Notion is "shots on goal," which encourages increasing the volume of experiments and new features. This approach recognizes that users are more likely to be upset by a lack of innovation than by minor accidental breaks, although a balance is crucial.

However, Max cautions against treating "feature count" as a primary metric, likening it to lines of code in its superficiality. He expresses a personal struggle with maintaining and improving software quality in the industry, even noting frequent regressions in current AI development tools.

Max emphasizes a desire for a return to "Apple-esque machined unibody aluminum kind of engineering" in software, highlighting a broader industry challenge to achieve high quality and reliability despite a focus on rapid iteration.

Shots on goal is a thing that we say internally a lot, which is like, "Great, how do you increase shots on goal?"

Developing Product Taste Through Iteration and Feedback

Taste in product development is the ability to accurately predict how a specific target audience, or "in-group," will react to a new idea or product. This involves recognizing what constitutes an "obviously good" solution, much like the initial reception of the iPhone or ChatGPT, which were immediately perceived as excellent.

Cultivating taste is not achieved through isolation, but by increasing "shots on goal" and rapidly iterating with feedback. This process is likened to training a machine learning model: inputs lead to reactions, and learning occurs through repeated exposure and adjustments. It takes considerable time and consistent effort, similar to a craftsperson refining their skill over many years.

Practical ways to build taste include constant exposure to other people's ideas, actively tinkering with new applications, and engaging in side projects to understand end-to-end product creation. Designers with high taste often take responsibility for entire products and continually experiment with new tools.

Surrounding oneself with examples of good design can also sharpen one's critical eye; for instance, naming conference rooms after iconic objects like the Macintosh or a Porsche 911 helps to benchmark one's own work against high standards. There's no shortcut to developing taste; it's a continuous process of learning and refinement through experience.

You're able to run a virtual machine in your head where given an idea, you can predict for a certain in-group whether they're going to like it or not.

Truly Great Products Are Built Around One Tiny Core Superpower

Truly successful products are not built by continuously adding more features, but by excelling at one exceptionally good

tiny core

superpower. This core is often discovered through a combination of luck and market validation, and it distinguishes great products from those that get trapped in a cycle of endless feature additions.

Examples of such

If I really look at the truly great products, they all have one tiny core that is so exceptionally good.

Applying the Jobs-to-be-Done framework for user-centric product development

The Jobs-to-be-Done (JTBD) framework prompts product teams to consider what users truly "hire" a product to do. It encourages an honest assessment of user needs, challenging creators to distinguish between what users genuinely want and what the product team wants users to want. This approach helps teams maintain a user-first perspective.

Applying JTBD allows product developers to zoom out and avoid getting too focused on the technicalities of building. It serves as a reminder to shift from an employee's perspective of "I made this thing" to a user's mindset, questioning if they would actually buy the product themselves. This often reveals disconnects between creation and user experience.

An example of applying JTBD is in crafting a landing page for a new feature. Teams often fall into marketing jargon, losing clarity. By asking how they would explain the feature to a friend or on a whiteboard, they can cut through the noise and ensure the communication directly addresses the user's specific "job" or need, making the product's value clear.

Ultimately, the framework ensures product teams don't lose sight of the core user problem they are solving. It's a tool to foster empathy and clarity, leading to products that genuinely meet market demands rather than internal aspirations.

The user hires you for a thing. Be that user for a second, would you even buy the thing that you just made?

Max Schoening's Hot Take: Knowledge Work Is Already Universal Basic Income

Max Schoening humorously suggests that universal basic income (UBI) already exists in the form of knowledge work. He posits that humans have created a hierarchy of jobs and tasks that go far beyond what's strictly necessary for basic survival and contentment. This extensive system of employment, particularly in knowledge-based roles, effectively provides a form of income and purpose to many.

He argues that regardless of AI's advancements, humans are inherently inventive and will always find new reasons and ways to insert themselves into productive loops. Even if the nature of these loops changes, the human drive to be part of the conversation and contribute will persist, leading to new forms of 'work.'

Schoening highlights the common observation that knowledge workers are often paid well 'just to sit in front of a computer and put the right sorts of words and letters into this thing.' He implies that while AI might challenge the specifics of these roles, the underlying human inclination to create and fulfill purpose through work will simply adapt.

my take is that we already have universal basic income, it's called knowledge work.

Max Schoening Shares Personal Insights on AGI, Contrarian Views on Exclusivity, and Lessons from Past Failures

Max Schoening discusses his approach to a hypothetical world with AGI where work is optional, stating he would continue tinkering and building. He views coding as an intellectual challenge akin to playing chess or Go, rather than solely a utility. This passion drives him to make the world more malleable through software, expressing openness to exploring robotics despite it being outside his current skill set.

He shares a contrarian perspective on exclusivity in product development, arguing that it's not always beneficial for products to be universally inclusive. Max believes in the power of small groups, suggesting that building a product tailored for a 'top of the class' audience, even if it excludes others, can be a winning strategy. He cites Notion as an example, positing that aiming for eight billion users might alienate its initial five hundred million core users, and highlights a podcast's success by deliberately targeting influential tech figures.

Max candidly discusses career failures, including an organizational mistake in hiring. He recounts how loosening the requirement for designers to code, initially done to ease hiring challenges, led to a loss of polymath skills within the team. He now believes it would have been better to have fewer, more versatile designers who possessed a broader range of abilities.

Another significant product failure involved his 2014 competitor to Notion. His team diligently focused on perfecting the editing experience with features like markdown folding, believing it was the crucial element. However, Notion, despite an initially 'terrible' editor, succeeded because its core block-based approach was the 'tiny core' that truly mattered. This experience taught him the critical lesson of identifying and building the correct core functionality, rather than spending extensive time polishing secondary features.

I think I am someone who I don't code because of a utility, I code because it's also an intellectual challenge. So I think of it as playing chess and go.

Max Schoening's Advice for Silicon Valley and Lightning Round Insights

Max Schoening advises younger individuals in Silicon Valley to resist the "frenetic" pursuit of trends and money, which often distracts from genuine passion. He suggests working hard, especially early in a career, but emphasizes the importance of prioritizing what truly matters over the fear of missing out or becoming part of a "permanent underclass."

During a lightning round, Schoening recommends several books, including "Code" for understanding computer fundamentals, "Tools of Conviviality" for contrasting human-centric tools with industrial ones, and "Seeing Like a State" which challenges executives to look beyond superficial systems to real-world team dynamics.

He shares his preference for entertaining movies like "Project Hail Mary" and notes the timely relevance of "The Handmaid's Tale" by considering AI as a parallel to the concept of God in the show. Schoening also highlights favorite products like the Ghosty Terminal Emulator, Moshi phone terminal, the open-source CORNE keyboard, and a high-quality Sevvy pocket knife, reflecting his focus on personal agency in computing.

Schoening's life motto centers on embracing uncertainty and the idea that "life is what you make it," encouraging people to enjoy the journey without seeking spoilers for the future. He concludes by urging listeners to recognize their innate agency, observing how everyday objects are built by people no smarter than themselves, and realizing they can learn to create most things from scratch within six to nine months.

just don't let the rush or the- Fren- frenzy sort of distract you from the things that you actually care about and are passionate in life.