Besides ignoring the alarmists in setting policy, what is really needed is commissioning "red team" climate models that do not begin with the assumption that atmospheric temperatures are primarily determined by greenhouse gas concentrations, and includes besides that, cloud formation, influences on cloud formation both human and natural, including cosmic ray flux, changes in albedo, urban heat island effect (dishonestly used by the alarmists to inflate the warming by favoring urban and suburban weather stations in their models), solar irradiance, natural cycles of oceanic heat-flux (e.g. El Nino and La Nina), and anything else that honest climatologists (there are some, e.g. Lindzen at MIT) think is relevant. The goal of the models should be to optimize fit to past data on several different time scale -- since 1970, since 1900, since 1850 (the one the alarmists favor), the past two millenia, the past 10,000 years. I strongly suspect the models resulting would give good justification for ignoring the alarmists, but might suggest suggest something useful by way of adaptation to the indisputable fact that the earth's climate does change.
