Nihilism, merging with technology, and why I still think we can fix the world

I hear a quiet nihilism in a lot of talk about AI. If minds are just machinery, and machines can now do much of what minds do, then why does anything we feel or choose matter? Add a news cycle about collapse and a feed that never ends, and "nothing matters" starts to sound like the grown-up position.

I don't buy it. I spend my days inside the machinery, and it has made me more hopeful, not less. Let me explain.

Being physical doesn't make life meaningless

Going from "we are physical" to "nothing matters" is a bad jump. Meaning was never floating outside physics, waiting for the universe to sign off on it.

When a child gets a proper meal at school, that matters because a real child is less hungry this afternoon and can actually focus in class. Nobody needs a cosmic stamp on that.

Nihilism gets one thing right: no meaning is written into the universe in advance. To me that isn't a tragedy. It's an invitation. We make meaning between people, in what we build and do for each other, and understanding the machinery better doesn't make that smaller. If anything, I feel it more. We're all part of the same flow.

Technology has already fixed a lot

"Technology will fix the world" sounds naive until you look at what's already happened. Since 1990, the share of people living in extreme poverty and the number of children who die before their fifth birthday have both fallen by more than half. Vaccines, cheap phones, mobile money and better crops did a huge part of that.

I see the small version of this in my own work. At Nourishubs, I run the technology behind a meal and nutrition platform. It started with school meals, and we've just opened it up to early learning centres and to organisations feeding their teams. Software gets the right meal ordered, paid for and delivered to the right person. Bienzoli, which I founded, gives small Mauritian shops invoicing and a point of sale on the phone they already own. Neither is glamorous. Both take friction out of someone's day, and that's the stuff I care about.

But technology doesn't fix anything on its own. It works when someone points it at a real problem, ships it, and stays accountable for what happens next. Without that, the same tools just amplify whatever was already broken.

We already merged

Merging with technology isn't science fiction. Your phone holds your memory. Your map is your sense of direction. Half your friendships live in your messages. AI as a thinking partner and brain computer interfaces are the next steps in a merge that's already well under way.

I want that merge. It's good when it extends what a person can know and do, and it goes wrong when it replaces your attention with someone else's goals.

What a collapsing model taught me

My research gave me the clearest picture of how this goes wrong. I ran a 3B language model as a "living brain" that remembered everything and learned from its own experience. Within a session, its speech collapsed into stock phrases like "im sorry" and "thank u".

The cause was a feedback loop. Its own degraded replies filled up the context it was imitating, so it kept learning from a worse and worse copy of itself.

People fall into the same loop. A feed that only shows you a slightly more extreme version of what you already think is that loop, running on a human. The fixes that worked for my model would work for us too: keep fresh input coming from the outside world, refuse to learn from junk, and check in on your health often.

Where I land

My optimism isn't a prediction. It's how I choose to work. The world gets better when people build useful things, measure whether they work, and fix them when they don't. It's slow and unglamorous, and it's still the most hopeful thing I know.

I'm 23, and I want to spend my life on that. If you're building something that takes friction out of people's lives, especially here in Mauritius or anywhere in Africa, talk to me.