Welcome back, friends.

Dr. Trump may have missed the mark…

An exhaustive review published in The BMJ finds no solid evidence linking paracetamol use during pregnancy to autism or ADHD. Previous studies hinted at a connection, but they failed to account for family genetics, parental health and other confounders [something other than the thing being studied]. When researchers controlled for those variables, the supposed risk disapeared.

This matters because paracetamol is the only pain reliever considered broadly safe in pregnancy, yet mounting scaremongering left expectant mothers stuck. The new analysis, covering 40 studies, concludes the quality of prior evidence was "low to critically low." Translation: ignore the shouty man. 

🧐 What's in it for me? If you're pregnant and in pain, the study suggests you can take paracetamol without spiralling into guilt. Regulatory agencies already said it was fine, but now the science backs them up properly.

💵 The Bet This won't move pharmaceutical markets, acetaminophen is off-patent and dirt cheap, but it could dampen the rise of "pregnancy-safe" wellness brands peddling dubious herbal alternatives. Companies like Tylenol (owned by Kenvue, spun out of Johnson & Johnson) benefit from renewed trust, though the real impact is regulatory. If anything, this underscores how prenatal health remains a minefield of bad science and litigation risk, which is why proper pharma stays cautious and generics dominate.

Travelling at Mach 10 makes the whole planet commutable. The snag is that commute involves temperatures that melt titanium and turbulence sufficient to redistribute your organs.

But maybe there’s a middle ground… A breakthrough from researchers at USC validates an old theory suggesting turbulence at Mach 6 behaves much like turbulence at slower speeds (Morkovin's hypothesis). Handy, as it means engineers wouldn’t need to reinvent aerodynamics from scratch.

The team used lasers to ionise krypton gas in a wind tunnel, creating glowing lines that twisted and bent through the airflow. After 11 years, they confirmed the turbulence patterns were extremely similar to those at lower speeds and this simplifies the design challenge enormously, making hypersonic passenger jets or space planes launched from runways far more feasible.

🧐 What's in it for me? If you're under 40, you might actually fly hypersonic before you retire. This could also transform access to space by making reusable air-breathing spacecraft practical, ditching rockets for runway launches.

💵 The Bet: If hypersonics work, aerospace giants like Boeing and Lockheed Martin may either look to acquire these startups or scramble to catch up. The real prize is whoever masters reusable hypersonic flight first.

We’re back to Photonics, not just because light powered computers are a sci-fi goldmine, but because the breakthroughs keep coming and the rate is increasing. 

Photonic chips promise to be faster and vastly more efficient than silicon but the problem has always been controlling microscopic beams of light without losing signal strength. Enter gyromorphs, a new class of materials discovered at NYU that block stray light from all angles, something existing materials could never manage.

Gyromorphs sit between the rigid structure of crystals and the randomness of liquids. They create barriers that stop light bleeding between signals, like soundproofing but for lightwaves. This could finally make photonic computing practical, routing light through chips without it degrading

🧐 What's in it for me? Photonic chips won't fit in your pocket anytime soon, but they could power the next generation of AI without burning through power grids. Expect data centres to get smaller, cooler and quieter. Then expect all those benefits to be traded in for having way more of them. 

💵 Out of the Lab: If gyromorphs deliver, they could slot into existing photonic architectures or spawn entirely new designs. The catch is manufacturing; photonic chips are still expensive to fabricate at scale. Whoever cracks cheap, high-volume photonic production could win the next decade of computing.

Your brain runs on 20 watts. ChatGPT's servers require a small power station. Aren’t we clever. Well, our window for smugness is closing.

Scientists at USC have now built artificial neurons that use silver ions to replicate how real brain cells work. With this, they’ve shrunk the neurons from hundreds of transistors down to one, creating fundamentally different engine that’s also a whole lot more efficient.

Traditional chips move electrons around which is fast but wasteful. The new "diffusive memristors" move silver atoms instead, mimicking the way calcium ions trigger signals in biological neurons. The result is hardware that learns the way a child does, i.e. from a few examples rather than thousands, and without the need of a nuclear plant's worth of power. If it scales, the energy savings could run to orders of magnitude.

🧐 What's in it for me? Phones might one soon be able to handle proper AI without melting your palm or needing constant cloud access. More immediately, data centres could stop being environmental catastrophes. The catch is silver doesn't play nicely with current chip factories, so don't bin your GPU just yet.

💵 The Bet: If neuromorphic chips deliver even half their promise, the semiconductor landscape fractures.

The real winner might be whoever cracks mass production first, whether that's a startup or a giant with deep pockets and the ability to manufacture. 

Wires are rubbish. When electricity flows through them, a substantial amount bleeds away as heat. Superconductors eliminate that waste entirely, but only when cooled to temperatures cold enough to liquefy air.

However, MIT physicists have now confirmed that graphene, when twisted at a "magic angle", superconducts through a completely different mechanism to conventional superconductors. In twisted graphene, instead of electrons pairing up through vibrations in the material's structure, they bond directly through their own interactions, forming weirdly tight pairs.

That distinction matters. Understanding why this mechanism works could guide the design of materials that superconduct at room temperature, which could mean lossless power transmission, compact fusion reactors, and quantum computers that don't require industrial freezers. The mechanism itself is still a mystery, but it's the clearest clue in decades.

🧐 What's in it for me? Don’t go stripping out your wiring just yet. Room-temperature superconductors are still a pipe dream, but if they arrive expect the world's electrical infrastructure to be in for a historic overhaul.

💵 The Bet: Hold your horses. The space is littered with hype and retracted papers. If graphene-based superconductors prove scalable, expect a gold rush, but scepticism is warranted until replication is ironclad.

Physicists claim to have mathematically disproven the simulation theory; the notion that we’re living inside a computer simulation created by a more advanced civilisation (think the Matrix, but peer reviewed). 

It turns out, it’s really hard to actually argue against the simulation theory but the new attempt hinges on Gödel's incompleteness theorem: any algorithmic system has fundamental limits, and since the universe apparently requires 'non-algorithmic understanding' at its deepest level, it can't be simulated.

It's an elegant claim but not watertight. The researchers argue reality emerges from an informational "Platonic realm" that underpins spacetime, and that this realm cannot be fully described by computation alone. Therefore, no simulation. The flaw? They're essentially saying "we don't fully understand physics, so it can't be computed," which isn't quite the same as proving it's impossible.

Until next time, stay curious.

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