Welcome back friends, and greetings to our new subscribers! 

Humanity's track record with 'unsinkable' isn't great, but researchers may have finally cracked it.

They created aluminum tubes with laser-etched interior surfaces that trap air, causing them to float regardless of damage, submersion time, or however many icebergs you throw at them.

The lasers create microscopic pits that make the surface superhydrophobic. When submerged, the tubes capture a stable air pocket that prevents flooding even when the structure is compromised. The team tested them in rough conditions for weeks, tilted them at severe angles, subjected them to “water impacts” (we assume they mean “waves”...), and punched in as many holes as they could. Regardless, it stayed afloat. 

They drew inspiration from diving bell spiders which carry air bubbles underwater, and fire ants which make floating rafts using their water-resistant bodies. Beyond unsinkable ships, this technology could also capture energy incredibly efficiently from moving water, think wave power 2.0.

🧐 What's in it for me? Applications will likely be mostly industrial in the short term: buoys, offshore platforms, emergency equipment, etc. But if this scales to shipbuilding, we’re likely to see a new era of maritime safety, or at least, a much more pleasant shipwreck experience. 

💵 Out of the Lab: Superhydrophobic coatings have promised much and delivered little, but this approach is different; the structure is the solution, not a spray-on layer that wears off.

After decades of dementia research and billions in funding, two new studies suggest some quick wins may have been in front of us the whole time. 

A JAMA study tracking 130,000 people across four decades found two to three cups of caffeinated coffee daily reduced dementia risk by 18%, with tea delivering similar benefits at one to two cups. Importantly, decaf did nothing. The effect held regardless of genetic predisposition to Alzheimer's which suggests this isn't just correlation dressed up by a coffee lobby. 

But even if you’re a caffeine starved Mormon, there’s still good news for you! A separate 20-year NIH-funded trial has produced what may be one of the most important findings in dementia prevention to date: only one type of brain training actually works. Speed processing exercises, where you quickly identify objects while managing distractions, reduced dementia risk by 25% in those who completed 14 to 23 hours of training over 3 years. 

Memory games and reasoning tasks, the foundation of most brain training apps, did precisely nothing. The theory is that speed training builds "cognitive reserve" by engaging broader neural networks, essentially creating backup circuitry for aging brains. A recent McGill imaging study confirmed the training actually reverses age-related decline in acetylcholine production, the neurotransmitter Alzheimer's patients are already short on.

🧐 What's in it for me? If you drink coffee, well done, keep it up. And if you really like your brain, speed training is available now on BrainHQ. This is the tool that generated the results, not a paid promotion (we don’t do that here), there may well be other companies that provide similar services. 

💵 Out of the Lab: Dementia prevention is becoming investable, and the winners may not just be pharma giants chasing amyloid plaques.

For the first time, AI has been definitively shown to be more creative than the average human. But before we spiral into existential despair, there's a catch.

The study pitted GPT-4, Claude, Gemini and more. against over 100,000 people on standardised creativity tests. The task: generate ten words as semantically unrelated as possible. It's called the Divergent Association Task, and it measures creative thinking by how far apart your word choices land in conceptual space. The models won, and it wasn't particularly close.

But it’s not quite that straightforward; the top 50% of human scorers still beat every AI tested, and among the top 10%, the gap widened considerably. Peak creativity, the kind that produces genuinely surprising connections, remains firmly human. At least for now…

🧐 What's in it for me? Current AI works best as a collaborator, expanding ideas rather than replacing them, so if you're highly creative, congratulations on being irreplaceable, at least for a few months. 

💵 Out of the Lab: If this creativity chasm persists (a big if considering the billions of dollars going into addressing this), we’ll likely remain in the "AI as tool, not replacement" world, which in turn could bode well for collaboration platforms over full automation.

Quantum computers need to be colder than deep space to work, but the cooling systems that get them there create enough noise to ruin everything. It's one of physics' more irritating ironies: the infrastructure required to maintain quantum states is itself a source of the decoherence that destroys them. 

However, researchers have now built a quantum "refrigerator" that uses noise to cool, creating the ironic equivalent of a double bluff. 

By injecting carefully tuned noise, the team can steer heat flow with extraordinary precision, measuring energy currents as small as attowatts (10⁻¹⁸ watts). To put that in perspective, warming a single drop of water by one degree at that power level would take longer than the universe has existed.

The device can operate as a refrigerator or a heat engine depending on configuration, which matters because quantum circuits will eventually need to manage their own temperature without million-dollar cooling infrastructure.

🧐 What's in it for me? Heat management is one of quantum computing's hardest remaining problems. Solving it inside the chip could accelerate everything from drug discovery to encryption. This likely won't affect your life this decade, but it might define the next one.

💵 Out of the Lab: The cryogenics market for quantum computing is small but growing fast.

First, jet lag may have finally met its match. A Japanese team identified Mic-628, a compound that advances the body clock by activating the Per1 gene. In mice mimicking a six-hour time zone jump, a single dose cut adjustment from seven to four days. Unlike melatonin, it works regardless of when you take it. Human trials are coming up soon.

Second, your sound machine might be sabotaging you. A Penn study found pink noise reduced REM sleep and increased nighttime waking. Earplugs worked far better. Given children spend more time in REM than adults, sound machines in nurseries deserve reconsideration. 

Finally, night owls face real consequences. A UK Biobank analysis of 300,000 adults found late sleepers were 79% more likely to have poor cardiovascular health. The silver lining: much of this was explained by fixable habits like diet and irregular sleep, not the chronotype itself.

Until next time, stay curious.

Like what you're reading? Share toast with a friend.