Science and health headlines can move faster than the evidence behind them.
A single study can become a confident headline. An early laboratory finding can become a wellness claim. An animal experiment can be presented as if it already applies to people. An observational association can be written up as if it proves cause and effect. A change in a biomarker can be made to sound like a meaningful clinical outcome.
That gap between what the evidence shows and what the headline implies is where deHype.io comes in.
deHype is opening its public beta to help readers check science and health headlines more clearly, more transparently, and with less reliance on instinct or brand trust.
The problem
Most people do not have the time to trace a health headline back through the source material.
A typical claim may pass through several layers before it reaches the public: a scientific paper, a press release, a journalist's summary, a social media post, and then further commentary. At each step, the wording can become more certain, more dramatic, or more commercially useful.
That does not always mean the article is wrong. Many science and health stories are based on real research. The problem is that the strength, limits, and relevance of that research can become blurred.
Important questions are often left unanswered:
- Was the study done in humans, animals, cells, or computer models?
- Was it a randomised trial, an observational study, or an early exploratory analysis?
- Did it measure a patient-centred outcome, or only a surrogate marker?
- Was the effect size meaningful?
- Were there important caveats?
- Does the headline match what the study actually found?
- Is the claim useful for real-world decisions?
These questions matter because health information can influence behaviour, spending, anxiety, treatment decisions, and public understanding.
What deHype does
deHype.io is designed to help users slow a claim down and inspect it.
Users paste a science or health article URL into deHype. The system then works back toward the source material and produces a structured evidence report.
A deHype report can include:
- The source chain behind the claim
- A summary of what the study actually did
- A claim audit comparing the headline with the evidence
- Caveats and limitations
- A clearer explanation of the likely relevance to people
- A grade or evidence signal showing how well the claim is supported
- Safer wording where the original claim appears overstated
The aim is not to replace expert judgement. It is to make the evidence trail easier to see.
A good health headline should not only be interesting. It should be proportionate to the evidence behind it.
What public beta means
deHype is now opening as a public beta.
That means the service is live enough for people to try, but still actively being improved. Some features, wording, report layouts, and evidence checks will continue to evolve as real users test the site.
During beta, the most valuable thing is feedback.
We want to know where reports are clear, where they are too cautious, where they are not cautious enough, where the wording could be simpler, and where the site does not behave as expected.
We are especially interested in feedback on:
- Accuracy of evidence interpretation
- Clarity of the report structure
- Missing caveats
- Confusing wording
- Broken or awkward user flows
- Topics users want covered
- Examples where the headline and evidence do not seem to match
Public beta is the point where deHype becomes more useful by being tested against real headlines, real readers, and real questions.
Who deHype is for
deHype is for anyone who wants to look beyond the first impression of a science or health claim.
That includes:
- Clinicians who want a quick structured view of a public-facing claim
- Researchers who care about how evidence is represented
- Journalists and editors checking whether a claim is proportionate
- Patients and the public trying to make sense of health stories
- Investors reviewing science-linked company claims
- Sceptical readers who want a clearer evidence trail
- Content creators who want to avoid overstating research
The site is particularly focused on health, medicine, nutrition, longevity, supplements, dementia, cancer, and other areas where public claims often run ahead of the evidence.
What users can try now
The simplest way to use deHype is to paste a health or science headline URL and generate a free evidence snapshot.
Try it with a headline about a supplement, diet claim, longevity finding, dementia risk, cancer research, or a new scientific discovery.
The question deHype asks is not just:
Is this true?
The better question is:
What does the evidence actually support?
That distinction is important. Many claims are not simply true or false. They may be promising but early, biologically plausible but unproven, statistically associated but not causal, or relevant only to a narrow population.
deHype is built to show that nuance in a clearer format.
What we want feedback on
During the beta period, we want users to challenge the site.
Please send feedback if:
- A report misses an important limitation
- A grade feels too harsh or too generous
- The explanation is too technical
- The explanation is too vague
- A source chain is incomplete
- The site does not handle a particular article well
- A topic area should be added
- A report would be more useful with a different layout
- The wording could be clearer for non-specialist readers
We also want examples of good test cases: headlines that are exaggerated, ambiguous, genuinely well reported, or difficult to classify.
The more varied the examples, the better the beta will become.
Why this matters
Science communication works best when it is interesting and accurate.
The public should be able to read health and science news without needing to choose between blind trust and blanket cynicism. Good evidence deserves attention. Weak evidence deserves caution. Early evidence deserves context.
deHype.io is being built to make that context easier to find.
It is not anti-science. It is anti-hype.