💰 markets: hype doesn’t scale, lessons from a $9B mirage

🧪 r&d: time as an afterimage, cosmology suggests that “time evolution” may not be a fundamental concept

🎨 creativity: quantum gaming goes public

🖋 notes on building: motion over perfection

📚 book club: what defines humanity? turtles all the way down.

📽 artifacts: the perfect organism

Within the sprawling terrain of deep tech, few stories are as cautionary (and yet insightful) as Theranos. Theranos remains the archetype of how hype can hollow out even the most seductive of visions. Once celebrated as a sleight-of-hand scientific miracle, the diagnostics company reached a $9 billion valuation by promising “revolutionary” blood tests using just a pinprick. But behind the curtain, the technology failed to work.

While this was a dazzling story of betrayal(recently inducted to the Netflix hall of fame), the collapse is interesting in that it revealed not only fraud, but a deeper systemic risk. When capital concentrates too heavily on a single story, the backlash can poison entire sectors, a fragility visible today across emerging tech.

Just last week, the Financial Times warned that the AI boom may be nearing a breaking point. Nvidia, now worth an astonishing $4.3 trillion, has helped lift global equities, but reliance on a handful of giants has created a precariously top-heavy market. Sam Altman himself admitted “investors are overexcited” and that many will lose money. Meanwhile, an MIT report found that 95% of AI investments are generating zero measurable return, which is quite a sobering reminder that commercializing frontier science is not instant magic.

The strain goes beyond public equities: nearly $3 trillion will be needed for AI infrastructure in the next three years, much of it funded by private credit and venture capital, creating new systemic risks if the cycle breaks. We also see extraordinary announcements of capital across quantum — Quantinuum, backed by Honeywell, is pursuing a fresh fundraising at a staggering $10 billion valuation, twice its worth in just 18 months, despite quantum computing still being nascent in commercialization.

Likewise, IonQ has captured attention far ahead of revenue-readiness, acquiring Oxford Ionics for over $1 billion. These valuations signal belief in sci-fi futures, but they also underscore the danger of letting investor sentiment overrun scientific patience and structural rigor.

To make the deep tech future not just dazzling, but durable, investment must merge ambition with discipline. How do we move forward?

By the Numbers:

Hype doesn’t scale. But patience pays.

The characters in the famous novel Flatland are bound to two-dimensional space, and we pity them (or perhaps envy them) for their simple lives, struggling only with two-dimensional problems. We imagine ourselves free in three dimensions, yet we bow to the tyranny of a fourth. But what if that tyranny is an illusion, if time is not the dictator we always thought it was?

What if time itself is not fundamental, but emergent?

A study from Princeton, the University of Chicago, and others introduces kinematic flow, a new way of describing the universe’s correlations without invoking time at all. Instead of tracking events through a ticking clock, they map the “geometry of interactions” in momentum space, watching how shapes and symmetries shift like patterns in a kaleidoscope. From these static boundary structures, the familiar arrow of time emerges as a kind of afterimage.

Quick Overview for Kinematic Flow and the Emergence of Time:

Time itself might be negotiable.

This past week at Gamescom, a game called Space Moths was generated live on a quantum computer. It isn’t the fault-tolerant machine we’re promised for the 2030s and it isn’t perfect. But for the first time, any player, passerby, or skeptic could sit down and interact with a quantum device through play. The team behind it, MOTH, is working on both pushing today’s hardware into creative use cases WHILE building the algorithms that may one day define fault-tolerant quantum computing.

Discovery is never clean. The first version of anything is by nature imperfect (the science of “the first pancake”). But existence is the threshold in that you can’t refine what isn’t yet real. Universum Labs’ MVP is meant to be the scaffolding of a resource, mapping problem → pattern → tool → application as a way of coaxing order from the noise of information. Reminders for you and me that clarity is not the prerequisite; it emerges after motion. Building, even in its messiest form, is the act of discovery itself. Prioritize being kinetic.

Re-reading Do Androids Dream of Electric Sheep (or the inspiration for Blade Runner” for all you movie > the book people) because the namesake behind the launch of this newsletter. An interesting exploration into what defines our humanity. Philip K. Dick framed it as empathy, measured and tested. But perhaps it’s something broader in the way we sense, adapt, and rearrange ourselves in response to the world. Machines may mirror intelligence, but it is our capacity to turn experience into meaning that keeps us on the human side of the line.

If you want to read along, you can find Do Androids Dream of Electric Sheep? here. (This is an affiliate link, which means I may earn a small commission if you purchase. Consider it a way to support Android Dreams while filling your bookshelf. Win-win.)