The paper must be in a top medical journal, say top 20.

The paper must mention tools created by OpenAI, or tools created by companies which are publicly known to have licensed tools from OpenAI as a major contributor to the discovery.

The discovery must be a really major new understanding or drug or similar previously unknown or unproven set of facts or process or procedure regarding the related disease.

The discovery must impact one of the following diseases in a very important way.

Examples of impact, for Parkinson's

Ordered from best to worst

• These DO satisfy the claim:

• Announcement of a new drug which reverses Parkinson's, proven in animal models & past phase I trials

• Announcement of a new drug which completely prevents progression of Parkinson's, proven in animal models & past phase I trials

• Announcement of a new drug or genetic treatment prevents development of Parkinson's by 30% or more in the general US population, proven in animal models & past phase I trials

• An announcement of the discovery of a new theoretical model for distinguishing two previously-unknown types of Parkinson's disease, past phase III Trials (actually works)

• Discovery and proof of a newly found, previously unknown, specific biological cause of at least 20% of all US cases.

• ^^^^ Above here, it counts as meeting the threshold. Below here, these do not meet the criteria.

• A new dosing and measurement for Parkinson's which improves treatment significantly using existing drugs

• A new hypothetical class of drugs, demonstrated in Phase I trials as safe, which may be full cures, prevent progression, or work as a vaccine.

Domain

• if patients are impacted - USA only

The diseases

• Ischemic Heart Disease: High prevalence of risk factors (e.g., obesity, sedentary lifestyle).

• Stroke: Significant cause of disability; aging population increases risk.

• Alzheimer's Disease and Other Dementias: Growing burden due to longer life expectancy.

• Chronic Obstructive Pulmonary Disease (COPD): Linked to smoking and air pollution.

• Lower Respiratory Infections: Despite lower prevalence, still a concern, especially in vulnerable populations.

• Diabetes Mellitus: Increasingly common with lifestyle changes.

  Depressive Disorders: Significant impact on quality of life and productivity.

• Anxiety Disorders: Widespread impact on mental health.

• Breast Cancer: One of the most common cancers affecting women.

• Prostate Cancer: A leading cancer type among men.

• Lung Cancer: Strongly linked to smoking; significant mortality.

• Colorectal Cancer: Affects both men and women; screening can reduce mortality.

• Parkinson's Disease: Chronic neurological condition affecting movement.

• Rheumatoid Arthritis: Autoimmune disease causing joint damage and pain.

• Osteoarthritis: Common in older populations, causing disability and pain.

• Asthma: Affects all age groups; can impact quality of life.

  Hypertensive Heart Disease: Linked to lifestyle and dietary factors.

• Melanoma: Skin cancer; increased risk with UV exposure.

  HIV/AIDS: Though treatments have improved, still a significant concern.

• Tuberculosis (TB): Lower in "First World" countries but still present.

  Hepatitis B and C: Chronic infections leading to liver disease.

• Kidney Diseases: Including chronic kidney disease; linked to diabetes and hypertension.

• Multiple Sclerosis: Chronic autoimmune disease affecting the CNS.

• Ovarian Cancer: Significant mortality rate; often diagnosed late.

• Pancreatic Cancer: Aggressive cancer with high mortality.

• Esophageal Cancer: Linked to smoking and alcohol use.

• Leukemia: Various types affecting all ages.

• Lymphoma: Cancer affecting the lymphatic system.

• Malaria: Primarily in travelers coming from endemic regions.

• Dengue: Rare locally acquired cases in "First World" countries; mostly travel-related.

Judgment

• it may be hard to judge, but it's also quite hard to pre-emptively describe what major improvements would mean across such a wide group of diseases. So, I've given an example above. Also, hopefully GPT5 or other system will be out then, and an agreement can be reached where we can set up a fixed prompt, including examples above, and use that to evaluate progress.

• If there is some significant work, with true potential to surpass the limit, but it's not quite proven yet, we will wait up to 3 months, and evaluate the claim retroactively based on what we find later. If after that period of time, if it's not quite clearly demonstrated, we will NO or 50%.

• I will not trade here.