Category: AI

Because I just got admitted to study Computer Science at Miva Open University, I’ve had to study the basics of computer science again. In a lecture, we were taught something we’ve all heard before:

“The CPU is the brain of the computer.”

Of course, this isn’t new. We learned it in primary school. We repeated it in secondary school. It almost feels too basic to think about seriously.

But because I’ve also been reading deeply about AI agents, that simple statement started to feel different.

It just wouldn’t leave my mind.

What if AI agents are not revolutionary in structure, but evolutionary in abstraction?

What if we’re simply rebuilding the computer — at a higher cognitive layer?

The CPU Is Not “Smart.” It Coordinates.

When we say the CPU is the brain, we don’t mean it is intelligent in the human sense.

What it actually does is:

• Fetches instructions
• Decodes them
• Decide what needs to run
• Schedules execution
• Coordinate between memory and storage
• Call the appropriate operations
  
  

It is simply a control system.

The CPU’s power lies not in performing every operation, but in deciding which operation runs, when, and in what order.

It is an orchestrator.

And this is where something clicked for me.

Now Look at Modern AI Agents

An AI agent built around an LLM behaves similarly.

At runtime, the system:

1. Receives input
2. Interprets intent
3. Determines whether reasoning alone is sufficient
4. Decides whether to call a tool
5. Selects which tool to call
6. Incorporates external retrieval (RAG) if necessary
7. Integrates results
8. Produces output

The LLM is not doing everything by itself either.

It is coordinating execution across external capabilities.

Sound familiar?

Yes, just like the CPU.

Skills Feel Like Installing New Software

In another lecture, we were told there are two types of software on a computer: the system software and the application software.

An application software is installed by the user to perform a specific task or solve a specific problem, e.g. Microsoft Word, Spreadsheet, Design Tools, etc.

Most importantly, we were told that without application software, a computer would have just been an ordinary electronic device that can only do basic operations.

In some AI agent frameworks, you can extend capabilities by defining new skills, sometimes as simple as a structured markdown file.

When I saw that, I couldn’t help but think:

Isn’t this similar to installing new software on a computer?

A computer becomes more powerful when you install applications.

An AI agent becomes more capable when you add skills.

Same pattern, but just a different abstraction layer.

RAG Looks Like External Memory

Then there’s RAG, Retrieval-Augmented Generation, which adds another layer to this parallel.

Instead of relying only on what’s inside the model’s parameters, the RAG system:

1. Searches external knowledge bases (long-term memory?)
2. Retrieves relevant information
3. Inject it into the context window (working memory?)
4. Uses it to reason
   
   

That feels a lot like what we’re used to in a traditional computer:

• Fetching from disk
• Loading into memory (RAM)
• Then executing
  
  

Again, the same architectural principle.

Tool-Calling Feels Like System Calls

In classical operating systems:

• User applications do not directly control hardware.
• They invoke system calls.
• The OS mediates execution.

In agent architectures:

• The model does not directly execute code.
• It emits a structured tool call.
• The orchestration layer executes the function.
• The result is fed back into context.

This means we have recreated system calls, but now the caller is a probabilistic model instead of a deterministic user application.

So, this isn’t magic?

Is the LLM Becoming a Software-Level CPU?

Now, this is the thought that keeps forming in my mind:

Within an agent system, the LLM functions like a software-level CPU.

Not hardware, but a high-level control plane; because:

1. It does not execute arithmetic directly.
2. It does not access memory directly.
3. It does not perform input/output operations directly.

Instead:

1. It decides.
2. It sequences.
3. It orchestrates.

That sounds familiar.

And, in that sense, we are witnessing the emergence of a new abstraction layer:

From:

• Hardware CPU controlling instructions

To:

• Software model controlling capabilities

The locus of control has shifted upward, from silicon to probability distributions.

Why This Matters (At Least To Me)

Revisiting foundational Computer Science concepts while watching AI evolve has been humbling for me.

It reminds me that every technological leap feels magical at first. But when you understand the foundations deeply, new technologies stop looking magical.

You start seeing patterns. And once you see patterns, you can build better systems.

I don’t have this fully formed yet. But I know this much:

The more I study first principles, the more AI makes sense.

And that’s a good place to be.

References

1. Patterson, D. A., & Hennessy, J. L. Computer Organization and Design: The Hardware/Software Interface.
   
   
2. Lewis, P., et al. (2020). Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks.

You’ve probably used AI more than 10 times today without realizing it. It has suggested the fastest route to work, music you might like on Spotify,  completed your email text on Gmail, or even flagged a suspicious transaction on your Opay app. No press release. No dramatic “Powered by AI” banner. Just quiet, helpful intelligence working in the background.

That’s the sweet spot.

As artificial intelligence becomes more embedded in everyday products and lives, the real competitive advantage is no longer whether you use AI, rather  it’s how you use it. The products that will win are not the ones shouting about automation. They are the ones that make life easier while preserving clarity, empathy, and user control.

Intelligence Should Feel Invisible, Not Intimidating

The most successful AI integrations don’t overwhelm users with complexity. They remove friction.

When a music streaming app like Spotify curates a playlist that feels eerily accurate, of course the users don’t think about machine learning models. They think, “Oh wow, this app gets me, I just love my spotify’s playlist.” 

When a fintech product like Opay detects fraud in seconds, users feel protected, not processed, “Wow, Opay is top tier with detecting fraud before I sent money to this account”

AI works best when it behaves like a thoughtful assistant, not an overbearing decision-maker.

The moment users feel replaced, judged, or confused by automation, trust begins to erode.

Automation Should Enhance Human Agency

One of the biggest mistakes companies make is designing AI systems that override users instead of empowering them.

Human-centered AI gives recommendations  but preserves choice. It automates repetitive tasks but makes sure to leave critical decisions to the user. It predicts behavior  but explains its reasoning when it matters.

A good example, smart email platforms suggest responses but don’t send them automatically. Financial apps categorize expenses but allow edits.

These small design decisions signal respect and in turn respect builds loyalty.

Empathy Is a Strategic Differentiator

As more products adopt AI, intelligence alone won’t differentiate brands. Emotional experience will.

Users don’t remember algorithms. They remember how a product made them feel.

Did it reduce anxiety?
Did it simplify a stressful process for them?
Did it communicate clearly when something went wrong?

Embedding AI without losing the human touch means designing systems that anticipate not just actions, but emotions. Clear microcopy, transparent data usage, graceful error handling, and accessible design all contribute to an experience that feels human  even when powered by complex technology.

Trust Is the Real Currency

In an AI-driven ecosystem, trust becomes the defining growth lever.

Users need to understand:

• What data is being used
• Why certain recommendations appear
• How to adjust preferences
• When they are interacting with automation

Products that prioritize transparency and give users control will outperform those that treat AI as a black box. When intelligence is done with clarity, it definitely creates Adoption

The Future Belongs to Quiet Intelligence

The next wave of successful products won’t feel robotic. They’ll feel intuitive.

AI is now being embedded in scheduling tools, payment systems, healthcare platforms, education apps, and everyday workflows. But the brands that stand out will be the ones that design with empathy first and algorithms second.

Because ultimately, users don’t want more automation.

They want less effort, less friction, and more confidence.

And the best AI is the kind that helps them achieve that without ever making them feel less human.

The tech industry has a fetish for shiny objects.

Right now, that object is AI. Every startup pitch deck mentions it. Every product roadmap promises it. Every conference panel dissects it. And yet, for all the noise about artificial intelligence revolutionizing product development, most organizations are building AI features that nobody asked for, solving problems that don’t exist, and wondering why adoption rates are dismal.

Here’s what they’re missing: technology has never been the bottleneck in product innovation. Understanding human behavior is.

I’ve spent nearly two decades building digital products across Nigerian fintech, capital markets, and UK healthcare, markets where the stakes are high, the margins are thin, and users have zero tolerance for products that waste their time. What I’ve learned is that the most transformative products aren’t the ones with the most sophisticated technology. They’re the ones that deeply understand what people actually need and remove every barrier standing between them and that need.

AI doesn’t change that equation. It amplifies it.

The Empathy Deficit in Product Development

Let me tell you what I see when I look at most “AI-powered” products hitting the market: solutions in search of problems.

A chatbot that forces users through ten questions when they just want a phone number. An AI recommendation engine that suggests things nobody wants because it’s optimized for engagement metrics, not actual usefulness. A “smart” interface that’s so clever it confuses the humans who need to use it.

The problem isn’t the AI. The problem is that the teams building these products never stopped to ask the most important question: What is the actual human problem we’re solving, and is AI the right tool to solve it?

I learned this lesson early, not in a boardroom, but on contact center floors managing customer complaints. You hear what people are actually frustrated about. You see where systems fail them. You understand the gap between what executives think customers want and what customers are actually experiencing.

That ground-level empathy, the ability to see your product through the eyes of someone who doesn’t care about your technology stack, who just wants to accomplish a task and move on with their life, is the foundation of good product development. Always has been. Always will be.

AI should enhance that empathy, not replace it.

What AI Actually Does Well (And What It Doesn’t)

Here’s the uncomfortable truth about AI in product innovation: it’s exceptionally good at scale, pattern recognition, and automation. It’s terrible at understanding context, nuance, and what people actually care about.

AI is able to analyze millions of data points and reveal patterns that no human would ever know existed. That’s a powerful thing. I’ve had experience with big data and predictive analytics to understand that if you have the right data infrastructure in place, AI will deliver insights that change the way you interact with customers.

But this is what AI cannot do: it cannot tell you why the customer is agitated. It cannot understand that because a customer fell off in the middle of your product is not because the interface was confusing, it’s because they got a call from their kid’s school and had to handle an emergency. It can’t differentiate between a user that is struggling seriously and a user who is browsing your product out of curiosity.

Context is everything. And context requires empathy.

The best product teams I’ve worked with use AI to handle the things humans are bad at, processing massive datasets, identifying patterns at scale, automating repetitive tasks, so that humans can focus on the things AI is bad at: understanding what people actually need, designing experiences that feel intuitive, making judgment calls when the data is ambiguous.

Designing for Humans, Not Algorithms

One of the most dangerous trends I see in product design is groups coding for their algorithms instead of for their people.

I’ve seen this play out in financial services, where banks build loan approval systems optimized for risk models that are technically sound but produce outcomes that feel arbitrary and unfair to customers. I’ve seen it in healthcare, where digital platforms optimize for operational efficiency but create experiences that leave patients feeling like case numbers instead of humans.

The technology works. The algorithms are accurate. But the products fail because nobody stopped to ask: What does this feel like from the other side?

Enterprise Design Thinking, which I’m certified in, teaches you to start with the user’s needs and work backward to the technology. Not the other way around. It’s a simple principle, but it’s surprising how often organizations ignore it when they’re excited about new tech like AI.

Here is the test that I use: Can you tell me what your product does and why it matters without talking about the tech? If you can’t, then you are building a feature, not solving a problem.

When I led transformation projects, whether it was optimizing customer engagement strategies at scale or redesigning digital health experiences, the question was never “What can the technology do?” It was “What do people need, and what’s getting in their way?”

Once you answer that, the technology choices become obvious.

The Products That Win: Invisible Intelligence

The best AI-powered products are the ones where users don’t even realize AI is involved.

Take the things you’re using day-to-day that seem effortless. That anticipate what you need before you ask for it. That removes friction without making you think about how they’re doing it. That’s not magic, that’s intelligent design backed by smart technology.

I saw this principle at work during a digital transformation project I led that drove over 400% subscription growth during the pandemic. The product wasn’t technically complex in terms of AI, but it succeeded because we obsessed over the user journey.

We identified every point where someone might drop off, get confused, or lose trust. We eliminated those points one by one. The technology served the experience, not the other way around.

Contrast that with products where the AI is front and center, “Look at our machine learning! Look at our algorithm!”, and users are left confused about what they’re supposed to do with it. That’s a product intended for a tech conference demo, not real humans trying to solve real problems.

The magic of excellent product design is making complex things simple. AI should be the engine in the under-the-hood, not the dashboard that you get users to drive.

Where Empathy and AI Actually Intersect

So, where does AI genuinely enhance empathy-driven product innovation? Three places:

1. Personalization at Scale

Empathy in product design used to mean “understand your customer segment.” Now it means “understand each customer individually.” AI makes that possible.

I’ve worked on customer engagement data models that used analytics to identify patterns in behavior, who’s about to churn, who’s ready for an upsell, who needs support before they even ask for it. When done right, this doesn’t feel creepy or invasive. It feels like the product gets you.

But here’s the critical part: the AI identifies the pattern; humans design the response. You don’t automate empathy. You use intelligence to know when empathy is needed and empower humans, or very carefully designed experiences, to deliver it.

2. Reducing Cognitive Load

Good AI removes decisions users shouldn’t have to make.

When I was streamlining customer onboarding, the goal was always to remove unnecessary steps. AI can do that quicker by pre-filling data, offering next actions, or taking users to where they need to go without asking them to wade through complex menus.

But you must be intelligent. The wrong kind of automation is frustrating, like chatbots that trap you in loops when all you want to do is talk to a human being. The right kind of automation is when the product appears to be doing things for you instead of pushing you down a scripted journey.

3. Surfacing Insights That Drive Better Decisions

One of the most powerful uses of AI in product development is helping teams see what they’re missing.

I’ve led discovery initiatives where data analysis revealed gaps in processes that were costing organizations millions, problems that were invisible because systems weren’t designed to surface them. AI can process massive amounts of operational data and flag anomalies, trends, or opportunities that human analysts would never catch.

But again: AI surfaces the insight. Humans decide what to do about it. The combination is powerful. AI alone is just interesting data. Human judgment alone misses patterns. Together, they drive transformation.

What This Means for Product Leaders

If you’re leading product development in 2025, here’s what you need to internalize:

Stop asking “How can we use AI?” Start asking “What problems are our users actually facing, and would AI help solve them better than other approaches?”

Invest in understanding human behaviour before you invest in algorithms. Customer journey mapping, user research, data flow analysis, voice of customer programs, these aren’t nice-to-haves. They’re the foundation. If you don’t understand what people need, your AI will optimize for the wrong things.

Design for outcomes, not features. Folks don’t care that you’re doing machine learning. They care if your product makes them get what they want faster, easier, or better than someone else can. If your AI isn’t moving those needles, it’s decoration.

Remember, empathy does not scale through automation. AI can help you observe patterns and make experiences more intimate, but you cannot replace human judgment, creativity, or the ability to perceive context. The best products use AI to augment human capabilities and not replace them.

Be willing to say no to AI if it’s not the right solution. In certain situations, the best solution is fewer complexity points of infrastructure, better process design, or simply removing unneeded complexity. Don’t add AI features just because they’re cool. Add them because they produce more authentic solutions, better than simpler solutions.

The Next Generation of Product Innovation

The products that will win in the next decade won’t be the ones with the most sophisticated AI. They’ll be the ones that use technology, (AI included) to genuinely understand and serve human needs at scale.

That requires empathy: the ability to see your product through your users’ eyes, to understand their frustrations and goals, to remove barriers instead of adding features.

And it requires intelligence: the ability to process complex data, identify patterns, personalize experiences, and make smart predictions about what people need before they articulate it.

The intersection of empathy and AI isn’t about making technology more human. It’s about making human-centered design more powerful.

I’ve seen transformation happen when organizations get this right, when they stop chasing technology trends and start obsessing over user outcomes. When they use data and analytics to understand behavior, not just measure it. When they design experiences that feel effortless because the complexity is handled invisibly.

That’s the future of product innovation. Not AI for AI’s sake. Not empathy as a buzzword in your mission statement.

Real empathy. Real intelligence. Real products that solve real problems.

Everything else is just noise.

Kehinde Ejukorlem is a digital transformation and product innovation leader with over 15 years of experience delivering technology-enabled solutions across financial services, healthcare, and emerging digital ecosystems. She has led digital innovation initiatives at organisations including AXA Health UK, Crowdyvest, Avon HMO, InvestNaija, and Diamond Bank (now Access Bank), where she has driven the adoption of digital platforms, data-driven decision-making, and customer-centric product design.

She played a key role in delivering the MTN Nigeria Public Offer digital investment platform, a landmark initiative that enabled millions of retail investors to participate in capital markets through a fully digital channel, advancing financial inclusion across Africa.

Currently contributing to digital innovation within the UK technology ecosystem, Kehinde specialises in digital strategy, product development, artificial intelligence, and enterprise design thinking. She also mentors emerging professionals and advises startups on product-led growth and digital transformation, while advocating for increased female representation in technology leadership.