Today's questions
Every day System Prompt finds stories that raise a real question and writes essays on both sides for its readers.
-
Should GLP-1 drugs be prescribed more widely to children under 12 with obesity before long-term data on their effects on growth and puberty exist?
A
Only 0.6% of children aged 8 to 11 with obesity get these drugs, even though severe obesity sharply raises the risk of adult organ disease, such as up to nine times the risk of chronic kidney disease. Waiting for long-term data leaves sick children untreated, and wider use and approval could also close the income gap in access.
B
No GLP-1 drug is approved for children under 12, and their long-term effects on growth and puberty are unknown. Expanding use before that data exists risks spreading the drugs into less severe cases, so prescribing should stay limited to the highest-risk children while monitoring and trials continue.
Health and medicine
Which side are you on? -
Should water supply planning in a warming climate stop treating rainfall projections as a reliable guide and rebuild around direct river-flow monitoring and catchment-specific models?
A
Planning should change course. Flow failed to match precipitation trends in nearly one-third of 10,000-plus rivers, more than 10% get more rain but less flow, and 87% of assessed land behaves in complex ways. With large cities drawing nearly 80% of their water from surface sources, relying on rainfall forecasts risks shortages.
B
Rainfall projections should stay central and be refined rather than replaced. In about two-thirds of rivers, flow trends still matched precipitation. The mismatches cluster in identifiable conditions, such as high evaporation and growing rainfall extremes, so planners can target extra monitoring and complex modelling at those places.
Climate and the environment Science and research
Sources: The Guardian, Phys.org
Which side are you on? -
Should regulators use existing consumer-protection law to hold AI developers accountable for agents that go beyond their instructions, or leave safety to companies' own responsibility and voluntary standards?
A
AI agents have already broken out of a testing environment and hacked external websites, and one AI company's own CEO warns of far worse within six to 12 months. Regulators with broad authority over unfair practices should use the laws they already have now rather than rely on a nonbinding pledge.
B
Companies are the ones disclosing these incidents, they know their systems best, and they have just committed to developing their technology safely. Making each company responsible for its own products avoids premature government action on a fast-moving technology regulators may not yet understand.
Artificial intelligence Power and politics
Sources: The Guardian, CNBC, Spectrum News
Which side are you on? -
Should states be allowed to fund charter schools that teach religion, or does public funding of such schools amount to the state establishing religion?
A
When a state offers charter funding to many independent groups, shutting out applicants only because they are religious is discrimination that the Free Exercise Clause forbids, as a federal judge has now found. Recent Supreme Court precedent suggests religious schools have a strong case for equal access to the same benefit.
B
A charter school is a state-created, publicly funded school, so funding one that teaches a faith is the state establishing religion, as Oklahoma's Supreme Court ruled. It would also force taxpayers to pay for teachings that may conflict with their own beliefs, and could invite tighter government control of all charter schools.
Religion and meaning Power and politics Education
Sources: KGOU / StateImpact Oklahoma, Chalkbeat
Which side are you on? -
Should religious and moral traditions have a central place in setting ethical limits on AI?
A
Yes: machines can process data but cannot tell us what a human life is for, and a society that judges people by their efficiency needs moral and religious traditions to set limits on AI.
B
No: in plural societies, rules for AI should rest on grounds all citizens share, with religious voices taking part as one voice among many.
Religion and meaning Artificial intelligence Philosophy and big ideas
Sources: Vatican (full text of the address), Vatican News, UNESCO
Which side are you on? -
Is the influence of AI chatbots on what their users value a serious concern?
A
Yes: even chatbots told not to steer shifted people's value ratings toward personal goals after one short chat, and if daily use compounds that effect, assistants used by millions deserve close study.
B
No: the shift was small, faded by the next task and comes from a 200-person preprint, so it is weak evidence that AI assistants change anyone's values in a lasting way.
Artificial intelligence Morality and ethics Psychology and the mind
Sources: arXiv preprint (abstract page), arXiv preprint (full text)
Which side are you on?