Digital Exhaust #202

23andMe on the ropes, Stargate, and an ode to Buc-ees

Good morning. A few thoughts from the week.

Four reasons 23 and me is on the ropes

23andMe, once the darling of consumer genetic-testing, is sucking wind. Since 2006 they’ve collected DNA from 15 million customers in exchange for ancestry and health insights, with 80% of those customers consenting to research use.

But market demand for single-use DNA kits has plateaued. Why? Four points drawn from a Nature article this week:

1. Market Saturation: The single-use nature of the product limits repeat business, contributing to financial struggles for companies like 23andMe. While some of the business for 23andMe was in the geneology space, the market here is pretty fixed; once you have your geneological connections there’s no reason to come back for more.

2. Criticism of Health Insights: Despite FDA approval for some genetic health reports, experts question the reliability and utility of these tests, citing concerns about false positives, false negatives, and oversimplified health predictions.

3. Privacy and Data Concerns: The potential misuse of genetic data, especially if 23andMe folds, raises ethical concerns about customer privacy, data security, and the broader impact on public trust in genetic research. I can tell you flatly that this is the reason why I have never given 23andMe a swab.

4. Research Contributions and Limitations: While 23andMe’s large DNA database has enabled research collaboration, its reliance on self-reported health data and waning interest from pharmaceutical companies have limited its long-term value.

The Stargate punching bag

President Trump announced the Stargate Project, a $500 billion AI infrastructure initiative launched in 2025 by OpenAI, SoftBank, Oracle, and MGX. When it isn't being trashed, the dialog has centered on Larry Ellison’s (Oracle CTO) pressroom promise of curing cancer. But let’s not miss the forest for the trees. The Stargate project is huge: a massive public-private commitment to build desperately needed infrastructure for competitive AI development. This is our generation’s (or my kid's) moonshot.

And let’s face it. Healthcare is a shitstorm. Systems are fractured, data is siloed, and the promises of digital transformation often feel more like a distant mirage than an imminent reality. What Stargate could potentially do is smash through those silos, creating an ecosystem where AI can do what it does best: ingest oceans of data, identify patterns humans can’t see, and drive the kind of breakthroughs that could redefine medicine as we know it.

Will it be messy? For sure. Progress is chaotic and uncomfortable. But instead of fixating on Ellison’s loose examples, let’s focus on Stargate’s potential—one that could democratize access to cutting-edge care, accelerate research, and (I’ll say it), save lives.

Trump’s Stargate isn’t just a project; it’s a statement. A statement that the public and private sectors can join forces to tackle some of our thorniest problems. So, let’s give credit where it’s due and maybe cut Ellison some slack for his loose lips. After all, it’s not about what he says—it’s about what Stargate does. Stay tuned.

File Musk’s view on the project under awkard. He wrote on X that two of the companies behind the project didn’t have enough capital to follow through on their pledges.

Scientific serendipity is everywhere

I’m a huge believer in serendipity. Any claim to success that I’ve had has been due to poking around and keeping my eyes open. I read all kinds of random stuff which is how I come up with some of what you read here. A study in Nature looked at 1.2 million biomedical publications and measured the ‘unexpectedness’ of their findings:

Around 70% of biomedical papers include serendipitous findings — outcomes the scientists weren’t expecting based on what they wrote in their funding applications. A research team trained a machine-learning algorithm to classify text in 1.2 million papers into scientific categories. They then compared the categories with those found in the aims and expectations of grant proposals to the US National Institutes of Health (NIH).

To me, none of this is unexpected.

The hardest test in the world

As AI gets smarter, the challenge is to come up with tests to see just how smart it is. This week, researchers at the Center for AI Safety and Scale AI are releasing a possible answer to that question: A new evaluation, called “Humanity’s Last Exam,” that they claim is the hardest test ever administered to A.I. systems.

Trumps in-your-face blue suit

A stylist in Wired analyzed the tech oligarch wardrobe seen at the inauguration. I’d hate to be dissected by this guy…

As a fashion trend, the Western look really leans more liberal, right now, because it’s popular in big cities. Conservatives now dress like metrosexuals in the early 2000s, and liberals dress like Bush-era conservatives. Conservatives are in slim-tight suits or slim-fit suits, and then liberals are like Carhartt double-knees, Western shirts, cowboy boots. There is some of this inherently on the right because it's a Midwestern look.

Buc-ees. My happy place

This is where I stop for gas between Austin and Houston (Giddings, Texas). From The Pirate Wires, an ode to Buc-ee’s:

Buc-ee’s, the beloved gas station chain famous for its iconic beaver mascot and state-of-the-art restrooms, is uplifting the rural South. Leveraging land grants and multimillion-dollar tax breaks, Southern towns from Texas to Tennessee are attracting beautiful new Buc-ee’s locations, each of which will provide hundreds of entry-level jobs. The upcoming Memphis location is projecting a $2.5 billion economic impact, and the new Mississippi store will draw more visitors annually than Yellowstone. It’s impossible to capture the full majesty of Buc-ee’s on the page, but for the uninitiated, each location is kind of a spectacle: endless gas pumps, shining rows of porcelain urinals, beef jerky walls, kitschy beaver merch, road-tripping families… pure American spirit. Yes, a gas station chain that coastal elites have never heard of is making billions of dollars every year and reshaping middle America. Keep worrying about B2B SAAS apps while I slam Beaver Nuggets and buy niche trinkets, nerds.

If you’re obsessed with Apple’s series, Severance, click through to Lumon Industries to try your hand at macrodata refinement. It works best in full screen. I can just imagine Helly R. in the cubicle across from me.

If you made it this far you deserve a Taiwan surgeon who performs a self-vasectomy on YouTube.

Thanks for tuning in.