I can’t believe that SXSW hits its forty years this year. Long enough for a scrappy and loud music festival to mutate into a city-wide experiment where startups, filmmakers, musicians, and optimistic innovation weirdos all try to prototype the future. It turned into one of the most complex and inspiring festivals on the planet.
I like to land in Austin a few days before SXSW detonates into its annual circus. It’s my yearly mental reset, a strange calibration ritual where culture, technology, and a mild amount of jetlag slowly align again. I need to feel the city before the crowd lands.
From my temporary resourcing den in Lakeway, perched above Lake Travis like a slightly caffeinated hermit, I watch Austin inhale before the storm. The early days have a specific flavor. Coffee, breakfast tacos, optimism, aggressive scheduling, and the faint smell of venture capital that has been slightly overcooked. The city sits there like a guitar on a stand, waiting for someone to hit the first chord. I realize -once again- how this festival is the metronome that beats the cadence of my thinking.
This year the circus lost its tent.
SXSW, the festival that began in 1987 and somehow turned Austin into an annual migration route for curious humans, turns forty this year. But the Austin Convention Center, the enormous brutalist, concrete hive that anchored the whole thing for decades, my home away from home, is gone. The city demolished it for a massive rebuild that won’t reopen until 2029. For a quarter of a century my SXSW map started there. Endless corridors, fluorescent meeting rooms, busy and inspiring blogger lounges, exhibition halls full of startups pitching the next revolution and quietly dying three booths later. Now it’s dust and construction fences. I had not foreseen that this deep pit would trigger so much emotions… I must be getting old.
In its place SXSW scatters itself across three “clubhouses” downtown. Innovation lives at Brazos Hall. Film and TV moved near the Paramount Theatre. Music landed in the Red River Cultural District. The rest spilled into hotel ballrooms across the Fairmont, the JW Marriott, the Thompson, the Hilton, The LINE, the Omni. I praise myself lucky I brought my most comfortable shoes.
Oddly enough, I think not having the convention center might actually make the festival healthier. Remove the giant convention box and the city itself becomes the infrastructure. Walking between sessions I pass street musicians, food trucks, overhear arguments that accidentally turn into new projects. Film people wander into AI panels. Startup founders stumble into music showcases. The collisions feel accidental in the best way. The theme this year, “All Together Now,” leans into that chaos. For the first time the three major tracks, Innovation, Film and TV, and Music, run simultaneously across the entire week: seven days, a plethora of things to do… one city-sized stage.
I warm up my brain at SXSW EDU, the education conference that quietly acts as the intellectual prelude to the larger festival. Education is where the real arguments live, and frankly where my heart is. Technology, policy, children, students, mentees and the future squeezed into the same rooms like a family dinner that is about to become interesting. The optimists show up. The cynics too.
This year the conversation has a new student in the room. Artificial intelligence walked into the classroom and pulled up a chair. Nobody fully understands this new arrival yet. Helpful, yes… but also slightly unsettling and suspiciously good at homework
One of the anchor keynotes comes from Adeel Khan, the founder and CEO of MagicSchool AI, a platform building AI tools specifically for teachers. His point lands with refreshing bluntness. Teachers are already experimenting with AI at scale. Some tools genuinely save time. Some reduce burnout. Some even improve learning. Others look like software written by people who have never seen the inside of a classroom. Khan repeats a line that spreads through the room like a small moment of relief: teachers are magic, not the AI. The machine can deal with the administrative sludge, grading drafts, organizing schedules, structuring lesson plans. The human does the part that matters: curiosity, encouragement… the ancient art of convincing a teenager that effort might actually pay off.
Another keynote pairs Bruce Reed, who leads AI strategy at Common Sense Media, the nonprofit that studies how technology affects children and was previously the White House deputy chief of staff under Biden and oversaw the administration’s AI strategy, with Laurie Santos, a Yale psychology professor best known for her course on happiness and her podcast The Happiness Lab. Their topic sounds polite enough: young minds and wellbeing in the AI era. But the underlying tension is obvious. Social media algorithms already shape how kids see themselves. Now generative AI enters the same ecosystem. The real question is not whether technology belongs in education, that debate expired somewhere around the second iPhone launch. The real question is what guardrails look like when the systems shaping childhood evolve faster than the adults responsible for them. “Guardrails”… I see it becomes one of those SXSW words that starts appearing everywhere.
Jennifer B. Wallace, journalist and author of the book Never Enough will talk about “mattering,” the deeply human need to feel valued and to add value. Simple idea. Big consequences. When people feel they matter, they show up differently. They build things, they contribute, they cooperate. When they don’t, you get loneliness, burnout, quiet quitting, or teenagers disappearing into algorithmic rabbit holes. I can feel something uncomfortable clicking into place in my head: we built systems optimized for performance. We forgot to check whether the people inside those systems still feel seen.
The Launch Startup Competition is where early-stage education startups get a small stage and a very sharp spotlight. The ideas here are still young, sometimes deliciously rough around the edges, but you can almost hear the gears turning and the possibilities popping.
Storytime AI from Baltimore experiments with generative storytelling to boost literacy, turning reading into something more interactive and personal for young learners. ShareTheBoard, a team out of Kraków in Poland, rethinks the humble classroom whiteboard, capturing lessons in ways that make them reusable, shareable, and less dependent on being physically present in the room.
Opal for Schools from New York focuses on digital wellbeing, trying to give schools tools to manage screen time and healthier tech habits. Youth for STEM Equity from Columbus, Ohio, working on something painfully obvious and historically neglected: expanding access to science and technology education for kids who were never invited into that pipeline in the first place. Rézme plays with the idea that the traditional résumé may simply be the wrong interface for a generation that learns through projects, communities, and nonlinear careers. It’s essentially compliance software aimed at fair-chance hiring.
And then there’s this years winner Apprentos, an operating system for apprenticeships, replacing pre‑internet bureaucracy with a modern stack, so employers can launch programs in minutes, governments can track ROI in real time, and talent gets mobile‑first career guidance rather than chasing paper trails. It sits exactly in that gap where real projects, real supervisors, compliance, and funding all have to line up before a learner ever gets near “meaningful work.”
Kudos to the jury. Different problems, different angles, same quiet rebellion against the idea that education has to stay the way it was…it mirrors my worries about educational immobility and unwillingness to change. Watching these founders pitch always resets my perspective. The room is full of slides, demos, nervous energy, and the occasional brave soul trying to explain a complex idea in three minutes without passing out. Some of the ideas will fail. A few will quietly disappear. That is the nature of early experiments. But every now and then you see the spark, the moment when someone describes a problem so clearly that the whole audience leans forward.
Education tracks (in schools, universities or companies) rarely change because a committee writes a perfect report. The systems are far too big and far too comfortable for that. Real shifts usually start smaller and messier. A teacher hacking together a tool. A founder frustrated with how learning actually works. A prototype that looks insignificant until suddenly thousands of people are using it. A supervisor fueling a mentee heaps of curiosity invoking suggestions.
Progress, in education as in most human systems, tends to arrive sideways.
For one strange week every year, Austin becomes the place where those sideways ideas bump into each other. In hallways, on sidewalks, in coffee lines, in bars where the conversation quietly shifts from music to machine learning to the future of classrooms. The city turns into a living laboratory of half-finished ideas and improbable collaborations.
Some of them will disappear by next SXSW. A few might change how the next generation learns. I for one look forward to that change, but I’m a bit weary as well. Whether we are ready or not, the AI governance (there, I used the word) guardrails are being built while the car is doing a hundred miles per hour down Congress Avenue.
If we don’t focus on the human connection -that spark between a mentor and a student that no Large Language Model can replicate- we’re just optimizing our own obsolescence.
