TLDR: Startup companies now charge up to $50,000 to predict embryo intelligence for IVF patients, but the science is shaky at best: polygenic scores explain only 4-6% of intelligence variation in embryos, with massive confidence intervals that could mislabel an average embryo as brilliant or vice versa. Real-world gains from selecting the "smartest" embryo average just 3.9 to 7 IQ points, while genetic trait selection risks unintended consequences like increased bipolar disorder risk due to gene pleiotropy. The practice raises urgent ethical concerns about wealth inequality and eugenics, is legal but unregulated in the US, and enjoys minimal public support—while experts argue we're better off investing in creating equitable environments where all children can flourish, rather than treating human potential as a customizable product.
You're scrolling through your feed and see it: a startup offering to predict your future baby's intelligence. For fifty thousand dollars. Not the IVF—that's extra. Just the intelligence report.
Your first thought is probably wait, what?
Welcome to 2025, where companies like Herasight, Nucleus Genomics, and Genomic Prediction are selling polygenic embryo screening to IVF patients who want more than a healthy baby—they want the statistically smartest one. Herasight's premium package promises unlimited screening for five years and an interactive widget where you toggle traits like you're building a custom avatar. Nucleus Genomics offers 900+ predictions for around $6,000. One company employee reportedly described IQ alongside other "naughty traits that everybody wants"—height, obesity risk, mental illness susceptibility.
The marketing copy sparkles with Silicon Valley optimism. But here's where curiosity needs to put on its skepticism hat.
Reading Tea Leaves in a DNA Smoothie
These companies use genome-wide association studies (GWAS for short). Researchers scan millions of genetic variants across thousands of people, looking for patterns that correlate with traits like height, disease risk, or cognitive ability. They distill these patterns into a polygenic score—a single number that sums up your genetic predisposition.
Think of it like a weather forecast based on DNA. Instead of predicting rain, they're predicting intelligence.
The catch? This forecast is blurry. Current polygenic scores explain about 16% of intelligence variation in adults. For embryos? That drops to 4-6%. Why? Because embryos haven't lived their lives yet. Intelligence is partly genetic, but it's also shaped by prenatal environment, childhood nutrition, education, sleep, trauma, lead exposure, access to books, and a thousand other things the score can't see.
Plus, you typically get 5-10 viable embryos per IVF cycle. That's not a vast menu—it's choosing between a handful of mystery cookies and hoping you pick the one with the most chocolate chips.
Reality Check: Confidence Intervals Run Wild
Here's where the hype wobbles. Let's say Herasight tells you Embryo A has a predicted IQ of 119. Sounds great, right? But the actual confidence interval might stretch from 98 to 143. That embryo could be average, could be brilliant, or somewhere in between. The prediction is less a bullseye and more a dart thrown while blindfolded.
Experts are waving red flags. A 2023 Nature study concluded these methods are "not sufficiently effective or robust" for embryo selection. The UK's Human Fertilisation and Embryology Authority echoes that evidence remains weak. The real-world gain? Selecting the "smartest" embryo from a typical batch yields an average boost of 3.9 to 7 IQ points. That's not nothing, but it's not Gattaca. It's more like a slightly better shot at the spelling bee, not a guaranteed Nobel Prize.
Then there's pleiotropy—the genetic version of "you can't have your cake and eat it too." Many genes influence multiple traits. Selecting for higher IQ might accidentally increase risk for bipolar disorder, shift height, or nudge obesity risk. A 2025 study examining over 1,000 traits found these overlaps are common and consequential. You're not editing a single gene—you're pulling a thread tangled throughout the genome.
When Enhancement Dreams Echo Eugenics Fears
If the science is murky, the ethics are a foghorn.
The $50,000 price tag means this isn't for everyone—it's for the wealthy. That risks exacerbating inequality in a way that echoes historical eugenics, just with a subscription plan and slick UI. The data also works best for people of European ancestry, because most GWAS research has focused on that population. We're looking at a future where advantage stacks on advantage, genetic disparity gets baked in, and we sort embryos like we're swiping on a very high-stakes dating app.
Public opinion reflects this discomfort. In 2025 surveys, 72-82% of people approved of using these screens to prevent diseases like diabetes. But for enhancing IQ? Only 30-36% said yes. We're fine with avoiding suffering; we're uneasy about manufacturing advantage.
When companies talk about ranking "naughty traits," they're commodifying human potential in a way that feels icky. It suggests embryos are products, and people who don't get selected are somehow lesser. That's not just a philosophical problem—it's a human one.
The Global Patchwork and What Comes Next
Here's the wild part: in the United States, this is legal and largely unregulated. The FDA hasn't stepped in. There's no official list of approved traits. Contrast that with the UK, where embryo testing is restricted to serious health conditions approved by regulators. Most of Europe is still debating frameworks.
Ethicists and scientists are calling for guardrails—voluntary moratoriums on heritable editing, transparent validation standards, broader societal conversations. The Polygenic Embryo ELSI Research Group is building frameworks to navigate these questions. The goal isn't to shut down innovation, but to steer it toward human flourishing rather than profit-driven enhancement races.
The Real Design Challenge
The desire to give your child every advantage isn't evil. Parents have always wanted better schools, better nutrition, better opportunities. The question is whether a $50,000 genetic prediction is the best way to express that love, or just the most expensive.
Intelligence isn't a single number you can preorder. It's a collaborative masterpiece between genes, environment, curiosity, struggle, mentorship, and chance. No algorithm predicts which kid will fall in love with astronomy because a teacher pointed out Saturn's rings. No polygenic score captures creativity that emerges from boredom, or resilience built from failure.
What we can do is demand transparency from these companies, push for regulations that prioritize equity, and remember that human potential isn't a product to optimize—it's a relationship to nurture. The most important traits we can select for are the ones no test can measure: kindness, wonder, perseverance, the ability to ask wait, what? when something doesn't smell right.
Maybe the real design challenge isn't choosing the "best" embryo. It's designing a society where every embryo gets a fair shot at becoming their own best self.