Why I’m rooting for AI, the biggest digital shovel since the internet

IntrodictIon

This isn’t just another “AI is coming for our jobs” or “AI will save the world” post. It’s a look at the plumbing of human progress.

If the internet was the highway system of the 21st century, AI is the fleet of autonomous vehicles, the logistics software, and the GPS combined. It is the “digital shovel” we’ve been waiting for—a tool that doesn’t just show us a pile of dirt (data) but actually helps us move it to build something useful.

There are two ways people talk about AI.

Some talk about it like it’s just a smarter Google: you type a question, you get an answer, and you move on with your day.

Others talk about it like a new programming substrate: something you can build on top of to create applications, automate workflows, write software, design products, run businesses, and accelerate research.

Whether you use AI as “Google with a brain” or as a platform for complex systems, there’s no denying one thing:

AI is the best thing to happen to humanity since the internet.

And yes, before anyone tells me I’m being insensitive, I sympathise deeply with those who have lost jobs to the efficiencies brought by AI. Me included. The displacement is real. The anxiety is rational. The unfairness is visible.

But if we zoom out, past the pain of transition, we can still be honest about the trajectory. I’m rooting for AI because it’s not a random tech fad. It is a logical progression of the digital age, and it solves problems that the internet exposed but never fully addressed.

To me, AI is the biggest digital shovel since the internet, because it doesn’t just give us information. It gives us the ability to move information, reshape it, apply it, combine it, and turn it into outcomes.

1) AI is the only logical progression after the internet

2) Knowledge-sharing is the only thing that moves humanity forward

If you study human history with a cold eye, you notice a pattern:

• Societies rise when they accumulate knowledge and pass it forward.
• Societies stagnate when knowledge is lost, hidden, or destroyed.

Wars, death, arson, conquest, censorship, and collapse have repeatedly erased libraries, destroyed records, silenced scholars, and wiped out technologies. Those losses didn’t just kill people. They killed progress.

We have lived through countless historical resets where humanity had to rediscover what was already known, because we didn’t preserve it, curate it, or distribute it widely enough.

Some societies are only now learning to do things that the Phoenicians were doing thousands of years ago, not because humans are incapable, but because knowledge wasn’t always durable, searchable, or shared.

Today, we’ve changed the storage part. We store everything.

But we’re still struggling with the curation part.

The amount of knowledge humans now generate and share is simply too large for manual indexing and searching. Even the best search engines are fundamentally matching keywords, links, and rankings, not understanding intent, context, meaning, or gaps.

AI changes that equation. It doesn’t replace knowledge. It organises it, compresses it, and makes it usable.

It makes knowledge liquid.

And when knowledge becomes liquid, able to flow easily across language barriers, educational barriers, and institutional barriers, humanity moves faster.

3) “How can I know what you know?” has haunted humanity for centuries

There’s a quiet question underneath almost every human frustration:

“How can I know what you know?”

It shows up everywhere:

• In school, when a student can’t access the teacher’s understanding.
• At work, when new employees can’t access the experience of seniors.
• In business, when founders can’t access legal, financial, or operational expertise.
• In healthcare, when patients can’t access medical knowledge in plain language.
• In communities, when people can’t access what others have learned the hard way.

A huge amount of human suffering is not caused by lack of intelligence. It’s caused by information asymmetry.

If I knew what you know, I wouldn’t worry about little things.
If I knew what you know, I’d avoid avoidable mistakes.
If I knew what you know, I could plan better, negotiate better, build better, and live better.

AI doesn’t magically eliminate inequality, but it attacks the monopoly of expertise.

For the first time, a person with curiosity and a phone can access structured guidance that previously required:

• a degree,
• a mentor,
• a network,
• or money.

That doesn’t mean AI is always right. It isn’t.
But the direction matters: it reduces the cost of accessing competence.

And that, on its own, is revolutionary.

4) Research: the world’s biggest digital shovel

This is where the “digital shovel” metaphor really earns its place.

Research has always been limited by human constraints:

• Time to read papers
• Time to synthesise findings
• Time to design experiments
• Time to analyse results
• Time to write and publish
• Time to replicate and validate

AI is accelerating each of those stages.

The question now shouldn’t be “Can AI do research?”
It should be:

Since we now have AI, the world’s biggest digital shovel, how many digital mountains are we prepared to move to level the ground for everyone?

Because this is the part people miss: the reason it feels like something is missing is because something is missing.

There are still many people who cannot meaningfully contribute to AI, either as builders, evaluators, domain experts, or data producers.

And the most important contribution to AI isn’t always writing models.

Often, the most important contribution is this:

Leaving a digital footprint for everything you do.

That’s how knowledge becomes trainable, searchable, reusable, and improvable.

But the digital divide prevents a large portion of the human race from doing that consistently, whether due to connectivity, costs, devices, infrastructure, language access, education, or political instability.

So yes, AI is a shovel. But not everyone has boots, gloves, or even access to the construction site.

If we want AI to accelerate humanity, not just a small slice of it, we have to treat digital inclusion as research infrastructure.

Because if people can’t contribute data, context, and feedback, then AI will keep over-representing the already digitised world.

What underdeveloped regions (especially Africa) can contribute to AI

When people talk about Africa in the context of AI, the conversation often becomes charity-flavoured: “How can AI help Africa?”

That’s the wrong framing.

A better framing is:

What can Africa contribute to AI, and what does AI miss without Africa?

We need to stop asking “How can AI help Africa?” and start asking “What can Africa contribute to AI?” If AI is to be truly “for humanity,” it cannot be trained solely on the data of the Western middle class. Africa, a continent with 1.4 billion people and a median age of roughly 19, represents the future of global data diversity.

Here are a few high-value contributions Africa (and other under-digitised regions) can uniquely provide:

1) Language diversity and linguistic innovation

Africa holds an extraordinary share of the world’s languages and dialects. Many are underrepresented in mainstream datasets. Building models that handle these languages isn’t just inclusion, it’s a technical frontier: code-switching, morphology, tonal languages, oral-first communication, and translation in low-resource settings.

If AI can truly serve humanity, it must understand humanity’s languages, not just the top 20.

2) Real-world constraint-driven innovation

Underdeveloped areas often operate under constraints that force smarter design: low bandwidth, intermittent electricity, shared devices, offline workflows, and cost sensitivity.

That environment produces innovations that wealthy regions ignore until they need them: efficiency, robustness, and minimalism.

As AI gets deployed “at the edge” (phones, local devices, factories, clinics), these constraints become global requirements, not regional exceptions.

3) New datasets from critical sectors

Agriculture, informal economies, cross-border trade, public health, climate adaptation, and mobile-first finance are not side stories, they are central to the future of the world.

Africa can contribute high-impact, high-variance data in areas where AI can deliver massive gains:

• crop disease detection and yield optimisation,
• logistics in chaotic infrastructure,
• epidemic surveillance,
• education at scale,
• identity and payments in mobile ecosystems.

4) Fresh research questions and evaluation benchmarks

AI is currently over-optimised for benchmarks that reflect the digitised West: curated tests, formal writing, predictable web data.

Africa can contribute what AI needs most: reality.

Benchmarks grounded in low-resource languages, noisy environments, limited context, informal text, voice-first interactions, and real operational constraints will make AI better for everyone.

5) Ethical leadership rooted in lived complexity

Many underdeveloped regions have long navigated high-stakes tradeoffs around governance, inequality, resource competition, and institutional trust.

That experience matters. It can shape AI governance around:

• data dignity,
• community consent,
• equitable benefit sharing,
• and accountability in environments where institutions are fragile.

In short: Africa isn’t just a recipient of AI.
It’s a missing pillar in AI’s claim to be “for humanity.”

The next steps in AI

If the last phase of AI felt like “chatbots,” the next steps are more structural and more consequential.

1) From chat to action: agentic systems

AI is moving from answering questions to executing workflows:

• booking,
• drafting,
• running analyses,
• monitoring systems,
• triaging incidents,
• generating and testing code,
• coordinating tools.

This is where productivity gains will compound, and where job displacement will accelerate unless societies adapt quickly.

2) Multimodal as the default

Text-only AI is the beginning. The world is not text-only.

The next era is models that understand and generate across:

• text,
• images,
• audio,
• video,
• documents,
• sensor data.

This unlocks huge value in healthcare, education, manufacturing, and infrastructure, especially where literacy barriers exist.

3) Smaller, cheaper, local models

Not every AI task needs a giant model in the cloud. We’re moving toward:

• on-device inference,
• private/local deployments,
• domain-specific models,
• efficient models that run under constraints.

This matters for cost, privacy, latency, sovereignty, and resilience, especially in regions with fragile connectivity.

4) Trust, provenance, and verification layers

As AI becomes widespread, society will demand mechanisms to answer:

• Where did this output come from?
• What sources were used?
• What uncertainty exists?
• What’s the audit trail?

The next step isn’t only smarter models, it’s systems that can be verified.

5) A new digital literacy: leaving footprints, curating knowledge

The most important “AI skill” won’t be prompting.

It will be:

• structuring data,
• documenting decisions,
• recording processes,
• creating reusable knowledge,
• building searchable institutional memory.

Individuals, companies, schools, and governments that master this will amplify themselves.

Those that don’t will be stuck reinventing wheels, while others automate bicycles.

The uncomfortable truth: I can root for AI and still demand a humane transition

I can root for the shovel while acknowledging that it’s going to be a messy construction site.

According to the World Economic Forum, 60% of workers in advanced economies may be impacted by AI by 2027. Some jobs will disappear; many more will be fundamentally transformed. Rooting for AI isn’t about being “pro-displacement”—it’s about recognizing that the “Information Overload” status quo is no longer sustainable.

The same way the internet created winners and losers, AI will too, but the difference is that AI is moving faster and touching deeper layers of work: writing, analysis, design, coding, planning, customer support, research, and decision-making.

So the real question is not whether AI should exist.

The real question is:

• Will we reskill at scale?
• Will we build digital infrastructure for the excluded?
• Will we redesign education for an AI-native world?
• Will we ensure benefits are distributed beyond a few companies and countries?

If we don’t, the shovel will dig deeper inequality instead of levelling the ground.

But if we do, if we treat AI as a civilisational tool rather than a tech product, then we might look back and say:

The internet connected humanity. We have a choice: use the shovel to dig deeper trenches of inequality, or use it to level the playing field for the billions of people currently excluded from the global knowledge economy.
AI helped humanity understand itself, and finally move forward together.