The End of Technically Optimistic — and What I’ve Learned

And four other lessons for today’s builder in my final post.

This is my last post here on Technically Optimistic.

I’ll share more soon about where I’ll be writing next, but for now: Thank you. Writing this newsletter, and making the podcast alongside it, has been one of the great joys of my time at Emerson Collective.

To those who read, listened, debated or sent me a note, it has been a true pleasure. You made this into a conversation, a community and a reminder that technology is never just about code and systems, but about people.

Over the years, we’ve heard so many voices together: Frances Haugen, Maria Ressa, Manu Chopra, Sneha Revanur and many more. I don’t name them as citations, but as companions and guides. Their words have shaped these lessons as much as my own.

So before I sign off, I want to leave you with five things I’ve learned. Five signposts for anyone building.

And the first is where it all begins, with the choice every builder has faced from the beginning: Whether our machines will replace us — or extend us.

Choose augmentation, not automation

In 1962, John McCarthy launched the Stanford Artificial Intelligence Lab (SAIL) to design systems that could stand in for human beings. Across campus, Doug Engelbart imagined tools that could amplify human intelligence instead. As John Markoff reminded us, that tension has been with us ever since.

Erik Brynjolfsson calls the danger of choosing wrong the “Turing trap”: Aim only to imitate humans, and wealth and power concentrate in a few hands. Aim to extend humans, and prosperity broadens.

That is the future worth building: one where machines don’t erase our role but expand it, freeing us to imagine more, create more, care more.

I deeply believe it’s possible. We’ve already seen glimpses of it. But don’t mistake this for inevitability. We only get that future if we actively choose it. And the choice is ours, right now.

Don’t just ship

I’ve said it many times: Launching a product is the beginning, not the end. The real work starts once it is in the world. Because if augmentation is the choice, then responsibility is the price.

We already know what happens when people forget this. Frances Haugen revealed how Meta’s own research showed algorithms nudging teenage girls from healthy recipes into eating disorder spirals. Arturo Bejar uncovered that one in eight teen girls reported unwanted sexual advances on Instagram in a single week. And Tristan Harris warned us that social media was our “first contact with AI,” optimized for addiction, unraveling trust before our eyes.

These weren’t accidents. They were reflections of choices — of what companies measured, rewarded…and ignored. Your tests reveal your values. And when you don’t test, that reveals your values, too.

We don’t need to keep relearning these lessons. We know the pitfalls too well. The task now is to think harder, not just faster. To bring diverse teams to the table, not just the usual voices. To ask not only “Can this work?,” but “Who might this harm, and how will we know?”

Testing is not optional. Harm is never hypothetical. Listen to whistleblowers. Watch what happens in the real world. Build with foresight, not just speed.

Because in the end, what you build shows the world what you truly believe.

Protect what makes us human

I once wrote about the student who, after speaking with a chatbot, told researcher D. Graham Burnett: “I don’t think anyone has ever paid such pure attention to me.” A beautiful and terrifying sentence. Machines can offer us infinite attention. But attention without accountability becomes a trap.

Rappler CEO Maria Ressa has shown us what happens when that trap is monetized, such as when Facebook built an empire on outrage and hate, eroding the very trust democracy depends on.

And memory is just as fragile. Datasets, archives and collective truths can disappear in an instant: a broken link, a vanished server, a government deciding that certain numbers are too inconvenient to keep online. Without redundancy — across geographies, institutions and communities — we risk not only losing information, but erasing history itself.

Thankfully, some are fighting to preserve it. The Internet Archive has saved more than 835 billion web pages and 44 million books and texts, a Library of Alexandria for the digital age. At the Harvard Library Innovation Lab, Jack Cushman and Jonathan Zittrain are mirroring government data before it can vanish, treating the public record as something alive, fragile, worth saving.

But it’s not only attention and memory we must guard. It’s also our creativity — our ability to imagine, to make meaning, to tell stories that move us forward. Justine Bateman warned us that reducing film and writing to “content” flattens art into something easy to automate, stripping away its soul. And Ari Melenciano has shown how AI can echo bias back at us, but also how artists can bend it, play with identity and stretch the ways we see ourselves.

Protect attention. Protect memory. Protect truth. Protect the human spark of creativity and imagination.

Attention is not infinite. Memory is not guaranteed. Truth is not self-preserving. Creativity is not inevitable. Defend them all as sacred. Build systems that outlast you. And make space for what only humans can create. Because someday, the future will depend on what we chose to save — and what we dared to imagine.

Center the vulnerable, seen and unseen

The strongest systems are the ones that begin with those most at risk. If technology works at the margins, it will work at the center. If it fails there, it will fail us all.

Rosalind Wiseman reminded us that young people are the subject-matter experts of their own lives. Sneha Revanur called her peers “the generation raised by AI,” the first to grow up inside algorithms, and the ones who will inherit their consequences.

We saw it in Manu Chopra’s work in rural India. Through Karya, he invited people, many of them women, speaking their mother tongues, to create data that trains the very systems shaping our future. In a single week, some earned more than their families had in an entire year. Manu calls it karya: Work that gives you dignity.

And Keolu Fox showed us what it looks like when Indigenous communities reclaim sovereignty over their own genomic data. For too long, their DNA has been mined for profit while offering little benefit in return. By building Indigenous-led biobanks, Keolu is proving that we can protect data, culture and identity at the same time.

It’s the same principle as the curb cut in the sidewalk: designed for wheelchair users, but ultimately useful for parents with strollers, travelers with luggage, kids on scooters and everyone else. Build for the most vulnerable, and you build for the many.

So put those voices at the center. Treat privacy, consent and dignity as non-negotiables. Listen to those on the margins not as edge cases, but as guides.

Build as citizens, not just coders

Building technology has never been neutral. Every choice is a civic choice, shaping who has power, who has a voice, who is seen and who is erased.

Maria Ressa said it best: Stop being just “users.” We are citizens.

And if that is true for all of us, it is especially true for builders. Every line of code is a cultural decision. Every model released is a kind of law. Every product is a piece of democracy, whether we admit it or not.

So don’t hide behind the fiction of neutrality. What you make carries your values. What you ship becomes our future.

Your code is culture. Your model is law. Remember that the future is not inevitable. It is being created every day, in your commit messages, your feature flags, your model weights, your data policies. Power rests with those who ship code. Build as if the future depends on it. Because it does.

May you choose augmentation over automation.
May you test for, and account for, harm.
May you protect attention, memory, truth and imagination.
May you build from the margins, centering the unseen.
May you build not just as coders, but as citizens.

And may the future thank you.