Will you become an AI shark or fairy?

…that’s the decision facing every company in the coming year.

One year into the AI arms race, the majority of corporations are in the same spot. They’ve spent the last year piloting use cases, focused on a few diddly internal tasks.

Most of these use cases are rather underwhelming; however, they’ve enabled companies to develop their governance models and technology platforms in a low-risk environment.

The turning point will come next. With a limited capacity, and a shortage on ML talent, every corporation will have to decide where to prioritize their AI/ML resources and investment.

The sharks will prioritize AI that automates parts of their business and reduces cost. These organizations smell the sweet, sweet efficiency gains in the water. And they’re salivating at AI’s promised ability to maintain productivity with less payroll (aka people).

The fairies will prioritize AI that magically transforms their products into something that is shockingly more valuable for customers. These organizations will leverage AI to break free from the sameness of today’s digital experiences–in order to drive lifetime value and market share.

Unfortunately for the fairy kingdom, most of the corporate business leaders I speak with don’t instinctively understand the value in the second approach.

However, they’re making a grave mistake: they will commoditize their product precisely when its potential value is exploding.

In the last few years, we’ve come closer and closer to the value ceiling of traditional digital products.

We’ve resorted to polishing every interaction, neurotically removing every speck of friction and trying to “delight” users with digital pleasantries. I know a lot of UI designers would argue there is more work to be done; however, from an innovation perspective, there hasn’t been much room to provide remarkably differentiated value to a user.

Don’t get me wrong, there is still bad UI out there, but the whole “we have an app that allows you to track your data in real-time, get push alerts, and take actions on your own accord” has become table stakes. It’s not moving markets anymore.

I believe our most important job, as Designers, is to create an AI product vision that illustrates how much differentiated value is on the table (especially for the companies that have done the hard work of building a strong UI foundation already).

The purpose of this is twofold: 1) to excite business leaders with the incredible opportunity to drive competitive advantage and 2) to make them fearfully aware of how risky it is not to pursue the fairy path.

So where will all this new value come from? In short, the transition from the software as a service (SaaS) paradigm to the partner as a product (PaaP) paradigm.

If you’re like me, you’ve probably labored your way through many Design Thinking workshops in the past decade, often culminating in solutions resembling a personal ‘concierge’, ‘butler’, or ‘coach’.

This solution archetype has always been logical because digital transformation stripped away the human partnership from our products, largely leaving users to fend for themselves. The allure of having a personal coach, assistant, or butler lies in the possibility of reintegrating that sense of partnership into products — at scale.

Yet, until the explosive and powerful dissemination of large language models (LLMs), we haven’t had the foundational technology to execute this in a meaningful way.

It’s becoming increasingly easy to envision a future where a company’s product is their AI partner, and where companies compete on their AI partner’s skill set rather than their software’s feature set.

My guess is that you’ve already been seeing glimpses into the partner-based future.

This is especially evident in moments like when OpenAI debuted its GPT Store. It was striking to see the store not filled with software applications, but teeming with thousands of AI partners, each ready to meet a wide range of nuanced needs — from comparing astrological charts to maintaining the health of fussy plants.

This vision is unfolding across major SaaS products, with the emergence of AI copilots, companions, assistants, navigators, and agents in tech products everywhere.

However, the ‘we have a chatbot too’ approach overlooks the profoundness of this new paradigm. The real value comes from the fact that computer programs don’t just execute functionality anymore; they partner with users to solve their needs.

This holds true whether the functionality is explicitly packaged and branded as a partner (like Microsoft Copilot, Salesforce Einstein, Zoom’s Assistant, Google’s Assistant) or more invisibly integrated (like Adobe Photoshop, Maze, and Tldraw).

GITHUB

Perhaps the most darling of all the copilots belongs to GitHub. As a SaaS product, GitHub allows developers to create, store, and manage their code, which was predominantly a solo endeavor until the arrival of this ‘pair programmer.’ GitHub Copilot now supports many core developer tasks, including asking questions about a codebase, writing code from a prompt, and suggesting code based on your repositories.

The impact of this new partner-based experience is staggering: it’s 55% faster to write code, 88% less frustrating, and approximately 61% of all Java code today is written with the assistance of Copilot.

With these outcomes, GitHub has transitioned into a new ‘Partner as a Product’ reality — where Copilot is valued as much as the SaaS product itself. This is explicitly stated on their website as “the competitive advantage that developers ask for by name.”

It’s easy to see how Copilot could evolve into an even more integral partner, further diminishing the relative importance of the interface it operates within. In fact, Cognition Labs is advancing beyond the Copilot model with its development of ‘the first AI software engineer,’ called Devin.

INTUIT

As a small business owner, I’m always digging around in Intuit’s Quickbooks product, which is basically accounting software for businesses that don’t have an accounting department.

I hadn’t heard anything of their AI vision, but it turns out they quietly introduced their GenAI-powered Financial Assistant late last year — which allows you to ask questions about business performance (rather than fiddling around in the bulky reporting section).

Am I mad about it? No. But the real value would come from growing Financial Assistant into more of a ‘bookkeeping partner’, supporting a user’s core needs.

Even after seven years of use, I still groan to myself every time I need to log into the interface. Being able to ask an AI bookkeeping partner my questions, send it my receipts, and trust it to catch errors would be so valuable — minimizing the role of the stocky SaaS interface.

MAZE

Maze offers researchers a SaaS platform to deploy user studies and capture valuable insights about their customers. The company has introduced AI features that assist in identifying biases in researchers’ questions, dynamically posing follow-up questions to respondents, and detecting themes in open-ended data.

Although these features are not yet integrated into a copilot format, they nevertheless lay the foundation for a future where ‘Partner as a Product’ becomes a reality. It’s easy to envision a fully realized research partner that figures out a researcher’s study goals and builds a custom study accordingly.

For instance, imagine a researcher looking to test three new product ideas but lacks experience in conducting concept tests. The partner could guide them through various methodologies, consider the researcher’s specific needs, and construct the study for the researcher.

In this envisioned future, Maze’s value proposition shifts towards the partnership it offers, rather than just the platform itself.

For me, these examples shift the reality that defines today’s products — and open my mind to a new future that’s not far off. Once you’ve achieved this perspective shift, the next step is to answer one key question: What does a partner-based future look like for my product?

In fact, I believe there is no more critical or impactful task for Design, especially at a moment when its value is being questioned. This is our moment to prove the indispensable role of Design in strategically shaping the future.

This vision of the future might appear self-evident to you. Yet, through 15 years of innovation work, I’ve learned to never underestimate the power of helping business leaders visualize a future that, to the design-minded, may seem blatantly obvious.

I outlined an approach for this at the end of my last article, but I’ll leave you with three criteria for a good vision:

• Holistic. Your vision should not just focus on ancillary benefits or functionality; it should visualize how an AI partner supports the core user experience.
• Vivid. Your vision should not just focus on the high-level shift to an AI partner; it should also vividly bring the future experience to life through tangible and recognizable moments.
• Realistic. Your vision should feel magically powerful, but it should not be fake or unrealistic. Ensure that you’re able to infuse a technical understanding of machine learning and GenAI to visualize achievable functionality.

This article was originally published in Empathy & AI, follow for more human-centered AI content or reach out on Linkedin.