A “confirmed” (or “proven”) intervention is one for which we have a lot of evidence. A “speculative” intervention, by contrast, is one for which we do not have a lot of evidence.
Global poverty interventions backed by randomized controlled trials are commonly viewed as confirmed, relative to interventions like value-spreading or improving others’ decision-making, which are commonly viewed as speculative. However, others argue that if we take long-term, indirect effects into account, then no intervention is confirmed in a strong sense.
Confirmed interventions are often preferred by those who are risk-averse, whereas those who simply seek to maximize expected value will treat confirmed and speculative interventions more symmetrically.
Dickens, Michael. 2015. Are GiveWell top charities too speculative?.
Hurford, Peter. 2013. Why I’m skeptical about unproven causes (and you should be too).