A screenshot of Basewell on desktop

A vision on the future of tech, AI, and human performance development

Our vision on the future of tech, AI, and human performance development.


Evan Stewart's headshot

Evan Stewart

July 31, 2023

7 min read

We believe the future of technology is adaptive.

That means tools will use AI to learn who you are and how you work, and will conform to meet you where you're at; automatically adjusting to your needs vs. forcing you to confirm to a rigid system.

This shift would be revolutionary.

  • Relationships between SaaS products and end-users would become symbiotic instead of linear.

  • A hyper-personalized UX would reduce friction (and reduce churn by association).

  • Companies would have more accurate data to solve increasingly complex problems.

The hard side of the problem is adaptive tools don't really exist yet. It's true terms like "adaptive" have been thrown around before (especially in the context of learning tools), but they've been primarily focused on content recommendations vs. understanding and conforming to user needs.

Netflix—while not a learning tool—is a prime example of this. You've probably heard Netflix uses an algorithm to change covers based on your viewing preferences? That form of "adaptability" is where most current learning tools end: surface-level.

In a more applicable example, say someone receives a low quiz score in a specific subject. Current platforms might make the following recommendation: train more in that subject until the quiz score is higher.

That's it.

Helpful? Maybe. Truly adaptive? No.

Critical context is missing to enable true adaptability. Yet unknowing consumers feel like things are adapting, because of that automatic content recommendation. It's a false-positive: a negative feedback loop that costs companies millions.

Wouldn't it make more sense to know if that person is:

  • Distracted, or disengaged? If so, why? Where?

  • Skipping content or just a fast reader?

  • Struggling with complexity, or just a slow reader?

  • Comprehending information or just good at guessing quiz questions?

Answering these questions should shift how the platform responds.

Knowing that a user struggles with completing complex, long-form, written content and has high distractibility in lessons more than 7 minutes is crucial. It's impossible to meet someone's needs without that info.

So, off-the-bat there's a LOT you can do when you start collecting that data. Yet, answering those questions is just one side to a more complex network of problems (more complex, in part, because humans are complicated).

The good news? While we're a complicated species, we're are also systematic and predictable. This is especially true in the context of corporate or educational environments, because training is standardized and outputs are measurable.

The question now is, how do you create software that's truly "adaptive"? By also learning from human interactions, at scale, across:

  • industries

  • geographies

  • workplace cultures

  • societal cultures

  • personalities

  • departments

and a myriad of other data that may influence how someone learns and produces.

Collecting all that info, you've begun answering an age-old question: "based on who [person] is: how they learn, where they work, and what they're expected to produce, how might their training experience be optimized to perfectly suit their unique circumstances?"

Current learning tools don't gather that info. Instead, they track unhelpful metrics like progression, recent quiz scores, and when content was completed.

How do you expect to fully understand your team's strengths and weaknesses if the context of who they are, what they've learned, and how they work, is missing?

Knowledge around how your team operates is your company’s most important data. Limitations to gathering those insights leave you with blind spots, waste, and in false-positive: a critically dangerous—sometimes lethal—combination. It's no wonder undertrained & disengaged employees cause $398B every year in waste.

With all that in mind, you might ask, "why does this problem still exist?"

Well, to be frank, the solution is really damn hard (not to mention out of scope for most companies). In order to start solving this problem, platforms must be equipped to collect, model, and train from critical customer data (surprise: current tools aren't engineered for this).

In addition, AI must be used for more than novelty. Building truly adaptive technology is more than streaming a return statement back from OpenAI's API. You must rigorously collect, build, and train (your own) models from (your own) proprietary information.

That's where Basewell comes in.

While we're providing best-in-class learning tools to seamlessly create, distribute, and measure training content, we're also collecting proprietary data to learn, predict, and model how people are solving problems on a global scale. True adaptability is born out of that effort.

Furthermore, true adaptability is impossible to achieve if you're not collecting that info to begin with. (That's one of the reasons for my belief that data is the ultimate moat of the next decade.)

Knowing how a:

  • [person] with [individual attributes]

  • working in an [industry] with [industry attributes]

  • utilizes [training] with [training attributes]

  • to produce [outcomes] measured in [specific ways]

on a global scale brings insights never available before. That level of intelligence allows you to do some incredible things:

1. Create a truly adaptable platform (one that's sensitive to the needs of—and reflexively adjusts to—each user).

2. Help companies better understand and support their people.

Using anonymized insights gathered across every interaction, Basewell AI will tailor the learning experience to each person.

That means your future employees never start from zero. 😉

3. Build the world's most powerful store of intelligence around workplace human performance development.

Knowing how factors across people, content, and organizations affect teams and outcomes carves a path to build optimized solutions to uniquely complex problems.

Learning tools are the perfect stage for this thesis, as transferring company knowledge is one of the most measurable rituals in the world.

All employees train and are expected to produce. Magic happens when you capture the context before, during, and after those learning moments.

We're building a future where tools can empower—not encumber—users. Where your world is optimized to meet your needs, and where your ability to learn and produce doesn't boil down to rigid, bloated, outdated standards.

Basewell is in beta, yet we're already serving incredible companies all around the world: capturing tens of thousands of monthly interactions (making content, training, reporting) from a wide-range of organizations from early-stage startups to enterprises with massive teams.

We're proving people are ready for the age of specialized, personalized tooling that's beautifully designed, doing obvious things well, and using AI in an approachable way (ie: under-the-hood vs. rubbing AI in your face).

My call to action is simple: if you believe in this vision, join us. Be a part of building the future. You can use Basewell right now to create, distribute, and measure beautiful training content. (Not to mention beta companies will be the first to access AI-powered adaptability when it launches).

If you have questions or want to talk shop, my DM's are open. For everything else, be sure to check out Basewell (or give Basewell a like here or on 𝕏/Twitter) to follow along.