1) We can be certain enough that we (mankind) are causing climate change that we should be obligated to act to address it. As CD says, 90% should do it.
2) While the cost studies are imprecise, there is a very real cost to unmitigated climate change that will likely dwarf expenditures we could take now to attempt to combat it. Among those costs are social instability and warfare that come from dramatically shifting availability of resources.
I just want to stop on the bolded part.
We should only be obligated to act to address it if #2 is true. Unless we can prove #2 at a high confidence level, and that it will dwarf expenditures we can take now to avoid it OR take down the road to mitigate it.
Which means we need to model a range of outcomes and determine probability. If we're talking about extinction-level or mass-starvation disruptions from climate change, then it justifies nearly any amount of effort to mitigate it. If we're talking about economic dislocation spread over the course of a century, but we can fundamentally continue feeding 7-10B people, sometimes you make a choice that the economic growth we generate now will effectively "pay for" the ability to mitigate things in the future.
People talk about all the amazing economic opportunities in green energy, but if we're shuttering useful energy production and claiming the great economic benefit of green energy,
it's just the broken window fallacy. That economic activity could go into other things to make our lives better.
I think that my confidence that CO2 is a greenhouse gas and that we're warming the earth, quite frankly, is a good measure ABOVE 90%.
It's when you start getting into trying to figure out how bad warming is, economically, and trying to compare that to the economic opportunity cost of focusing on incentivizing green energy over what the world could otherwise inventing to improve the world? I'm not sure I know the answers there.
Therefore I can't say that because we're causing warming, we should be obligated to stop.