Somewhere tonight, someone is up. Phone brightness turned all the way down so it doesn’t wake anyone, trying to understand a diagnosis that just landed on her family. She finds the paper that might actually answer her question. Then she hits a wall: $39.95 to read it, no way to know from the abstract alone whether it will even be relevant. She pays it anyway, because what else do you do in that situation?
This is a post about Open Access: the simple, embarrassingly obvious idea that research paid for with public money should be free for the public to read. It’s not a new idea. What’s new is why it matters more right now than it ever has.
AI has changed the stakes. Tools now exist that can search real scientific literature and hand you a plain-language answer in minutes, which means more people than ever are trying to access the actual research, not just a headline about it. Every one of them runs into the same wall the person above did. Open Access isn’t a nice-to-have anymore. It’s the thing standing between good information and slop.
This is also a good moment to say why this post exists right now. The Mind pillar just wrapped, the longest series this newsletter has run so far. Before starting the next one, a reader asked a question that deserved a proper answer rather than a quick reply in the comments: how I actually build the research and the framework behind this newsletter.
That felt worth pausing for. So, three posts, a short intermission before the blueprint continues: this one on finding scientific publications, next week on judging what you find once you have it, and the third answering that reader directly. Then it's back to the pillars, starting with the Brain area of the Body pillar (a brand new addition!).
New here? Start here:
Swiss Army Mum is a practical guide to long-term health for busy women, built on four pillars: Body, Mind, Glow, and Flow.
Not every tool. Just the right ones.
Open Science Is Simply Science Done Right
I’ve spent the last nine years of my professional life on exactly this problem: encouraging researchers to publish openly, helping them do it properly, and measuring what actually happens when they do. This is called Open Access, research that’s libre1 for anyone, not just people at institutions that can afford a subscription. It’s one piece of the broader Open Science movement, and it isn’t a hobby I picked up alongside the newsletter. It was the day job. Nine years in, I’ve landed on a fairly blunt conclusion. Open Science is simply science done right.
In the conventional publishing route, scientific journals acquire free content (researchers give away their papers - and their copyright!) plus free labor (peer review, mostly unpaid) and sell them back to institutions at high margin.
Yes, you read right.
A large proportion of scientific research is funded by public money: government grants, university funding, taxpayer-backed institutions pay for the research and the researchers who carry out the studies. They then submit the work to journals, who send it out to reviewers (them too, mostly paid with public funding) who check the quality of the paper and give feedback (for free!) to the journal. The research is ultimately published behind a paywall. Public money is used again, through the institution’s libraries to pay access to the research it produced. Private individuals are of course cut out of access.
It’s actually a brilliant business model, with astronomical profit margins, built on public funding and fueled by journal prestige and marketing.
This led to what was known as “the serials crisis” which denotes the rise of journal subscription costs far faster than inflation and library budgets.
Open Access is the movement to counteract just that. Born in the early 2000s, with the rise of the Internet, the rationale was that producing and sharing scientific papers was easier, cheaper and faster than ever. There was therefore no good reason for subscription costs to keep increasing and all the reasons to share the publications with those who financed them (the tax payers!).
It goes even further than that. It's written into the Universal Declaration of Human Rights2. Article 27 states plainly that everyone has the right "to share in scientific advancement and its benefits." Nobody enforces this in any meaningful way, which is sort of the whole problem, but it's worth knowing the case for open access isn't just mine. It’s, at least on paper, everyone's.
Publicly Funded Should Mean Publicly Available
Publicly funded research should be publicly available. That isn’t a radical position. It’s just the plain conclusion of who paid for it in the first place and the advancement of human kind.
Why Open Access Matters Now More Than Ever
Here’s the twist that makes this urgent rather than merely annoying. AI-powered literature search tools mean a curious layperson can now do, in an afternoon, roughly what took me years of graduate training to learn to do by hand: search the primary literature, follow citation trails, and start to build a real picture of what the evidence actually says.
That’s a genuinely good development. But it runs straight into the paywall problem, at exactly the moment more people are hitting it directly rather than never trying at all.
Think of it in the opposite direction. What an AI tool can retrieve and summarise for you is shaped by what’s actually accessible to it, the same way it’s shaped by what’s accessible to you.
Lock research behind a wall and you don’t just block a stressed reader in the middle of the night. You block the tool she’s using to help her. What’s locked away doesn’t just fail human readers, increasingly, it’s invisible to the AI tools people now ask first. If we want reliable information coming from AI, it needs to be fed reliable information.
Not Just in The Middle of The Night
That middle-of-the-night scene is the most urgent version of this problem, but it’s far from the only one. Most people reading this newsletter aren’t searching for a loved one’s diagnosis tonight. They’re just curious, a little skeptical of whatever wellness claim is currently trending, and they want to know what the evidence actually says before they act on it: on seed oils, on magnesium, on whatever the algorithm insists will fix your sleep this week. That’s a completely legitimate reason to want real research too, not just a crisis.
This is where AI changes the equation for that reader as well, not only the one in crisis. A few years ago, checking a claim yourself meant trusting whoever posted it, or spending an afternoon you didn’t have digging through a university database. Now you can ask a tool to search the actual literature and hand it back to you in plain language, in minutes. For a naturally curious, critically minded reader, that’s not a small thing. It’s the difference between accepting a claim on faith and actually being able to check it yourself.
One important caveat before you go do exactly that. These tools can still get it wrong. AI summaries occasionally misread a study, overstate a finding, straight-up make up a finding or, in some cases, describe a source as saying something it doesn’t actually say, a failure mode generally called hallucination. The fix isn’t to distrust the tool completely. It’s to treat its answer as a strong first draft, then click through to the actual study before you build a decision on it. Ask the tool, then verify the source.
AI Summaries Are a First Draft
AI research tools are remarkable for finding and summarising real evidence quickly. They still make mistakes. Treat every AI summary as a first draft, and check the source before you trust the conclusion.
Why Every SAM Post Stays Free
This is also why every single post here is free, and always will be. Not because free is some clever growth hack, though I won’t pretend it doesn’t help. Because after nine years watching what happens when good research sits behind a wall nobody but a university library can afford, I’m not about to build a smaller, cuter version of the business model I just a few paragraphs tearing apart.
The science and the frameworks stay free forever. If a paid tier ever adds anything, it’ll be the toolbox, not the information. That’s a line I’m not moving, and if I ever do, you have my full permission to remind me of this paragraph.
The Tools Actually Worth Bookmarking
Two tools are worth your time right now if you want to search real scientific literature the way I do.
Consensus3 is the one I keep coming back to, and it deserves more than a passing mention. Ask it a real question, “is the HPV vaccine safe?”, “does magnesium help with sleep?”, and it doesn’t just hand you a list of links. It gives you an overview: a plain-language synthesis of what the collected research actually says, plus a visual meter showing how much agreement exists across studies, yes, no, or somewhere in between. From there you can filter results by study type, narrowing down to just meta-analyses or just randomized controlled trials, and by subject, so you’re not wading through results that have drifted into a tangentially related field. That filtering is worth its own moment. Next week’s post, on actually grading evidence, will walk through exactly how to use those filters to separate strong research from noise, so think of this as the setup and that one as the demo. It also has a free tier, so there’s no reason not to try it tonight.
Semantic Scholar4 is completely free, needs no account, and is the best place to actually go find the underlying paper once you know roughly what you’re looking for. It’s more of a publication database. Less powerful if you’re not comfortable around those.
One more worth knowing about, even though you most likely can’t use it freely. OpenEvidence5 is arguably the best clinical research tool that exists right now. Its free and unlimited use is also restricted to verified US healthcare professionals only.
Finding the research is half the skill. The other half is knowing what to do with it once it's in front of you, because not every study that turns up deserves the same trust.
Next week, I'll walk through the exact framework I use to research any claim: the handful of terms that decide how much weight a study can carry, a ladder that ranks evidence from weakest to strongest, and one real example, the claims about creatine that are everywhere right now, graded from start to finish.
See you next time!
This is educational, not personal medical advice. Your biology, history, and context matter. Work with a qualified healthcare professional.
References
Libre means free to read and also free to reuse, adapt, and redistribute, usually under an open license like Creative Commons. It's a stronger standard than merely "gratis," which just means free to read (as in without cost) with no reuse rights attached. Most of what gets called Open Access today is gratis only.