the arrogance of the privileged

NASA admin Mike Griffin noted that deciding the current climate is the best is a rather arrogant position.

Ok, there is a point there. But, we're making a choice whether we like it or not, so what should we choose?.

Not choosing is also a choice, and one no less arrogant.

The preindustrial mean atmospheric carbon dioxide concentration was about 275 ppm, plus or minus maybe 5%, with rare intermittent spikes which drew down rapidly, probably due to volcanic injections followed by oceanic and biospheric re-equilibrium.

The current concentration just went over 375 ppm and is trending inexorably up.
The current increase is anthropic, due to burning of fossil fuels, currently oil is a significant contributor but in the long run it is really the coal that matters, that is where a lot of buried accessible carbon is, along with the temptation to use it.

We're headed for doubling of CO₂ sometime later this century - exactly when depends on demographics, economic trends and technology development.
The crude mean impact on the global climate due to this change is a small change in the mid/near infra-red atmospheric opacity, which will lead to a small fractional change in the mean effective temperature of the Earth. To order 1.5 (direct forcing plus water vapour feedback) this will be about a 1% effect, or a change of about 2.5 K in mean temperature. Empirical data suggests the net second order effects will be slightly positive and the net response will be more like 3-3.5 K.
More interestingly, the effect is heterogenous (eg night time vs day time warming and pole vs tropical warming); even more interestingly the there will be strong but poorly quantified perturbation on secondary climate indicators - like precipitation patterns.

If we extrapolate further, we're headed for ~ 800 ppm CO₂ in the near historical future (going higher relies on very unreliable estimates of future coal recovery). That is getting to the "you have got to be kidding me, oh my god we're all going to die" territory, with significant qualitative climate change to states not seen for tens of millions of years.

Seriously, it would cause major climate shifts, likely significant macrofaunal extinctions, probably major human impacs (ie LOTs of people dying) and major economic impacts.

We can stop this.
I don't think it is feasible to stop short of 400 ppm, unless there is a catastrophic human dieback in the near future. Doubling is probably inevitable.
Natural drawdown timescales seem to be of order 100,000 years, from the record of CO₂ spikes in the last 30 million years or so.
We could make a serious effort to draw down the peak CO₂ ourselves.

BUT, whaterver we do, we are making a choice, and it is an arrogant choice no matter what.

Yes, slamming the brakes on fossil fuel burning as a matter of overriding priority is a choice, an arrogant choice, and one which "decides for others" what is best.
But, so is the status quo - every gram of fossil fuel burned is an implicit, and very arrogant choice about what is best, not just for you, but for everyone else.
And this is a regime where microeconomics does badly - the marginal cost to me of filling up the car or leaving the light on does not reflect the medium term externalities, and not at all the long term risk of catastrophe.

So... what do we do?

Well, we can turn the question around: where do we want to stop?
Drawdown is possible but slow, and doing it too rapidly could be as catastropic as the rapid increase; a big part of the hazard is not just change, but rapid change.

So, the peak matters (we're going to hit the peak fast compared to any relevant timescale).

Now, I don't think we want to go back to the Eocene - just too different, too wild.
Older climates than the eocene are poor indicators, because the continental topography was too different and at some point the Sun signifcantly cooler.

But, do we really want to go back to 275 ppm or lower? I'm not sure the deep Pleistocene glaciations were that interesting.

And the miocene was just too boring

We can go back to the Pliocene
but where we are headed is the mid-Oligocene

Is that what we want? And who are "we" anyway? Suburban americanites? Rural bangladeshis? Icelandic sheep farmers?
Even locally there might be some dissent - the southwest US residents might view the prospects summer highs of 50+C and decadal droughts somewhat differently from northeasterners contemplating mild winter nights and Virginia-like winters; or is it increased precipitation and small temperature changes in the southwest, combined with hurricane tracks and Florida like summers in the northeast...?

Part of our problem is that we do not have enough information to decide in detail. A good initial approach to that problem, one might think, would be to gather as much information as possible (as opposed to, you know, cutting down on the information gathering...).
Secondly, we can think about slowing down the process as much as "feasible" - to have more time to make a rational decision.
What we can not really do anymore is continue as we have, and the real problem will be coal, we have to stop burning it on a time scale of a small number of decades (or at least cut back by at least an order of magnitude) or we will have made the most arrogant decision of all.

There are two additional issues - one is that we can change our minds, for all practical purposes, most of our decisions are non-binding on the future generations. Except in so far as some choices we make can eliminate the range of choices they have. It is possible that 60 years from now a deliberate choice will be made to slowly ramp up CO₂ concentrations to 550 +/- 20 ppm and leave them there for a millennium or ten, but if we have already forced it up to 800 ppm by then, that will not be much of a choice.
There is a second trap there, which is that we can not incur an excessive opportunity cost to preserve the future choices - crudely put, if we freeze to death because we wouldn't burn coal, then we won't have grandchildren to make choices about whether to burn the coal we left.
But, the opportunity cost evaluation must be weighted by the risks and uncertainties - economists tend to assume zero or narrow spreads in long term risks and focus on the short term costs; while environmentalists tend to focus on poorly quantified long term risks to the exclusion of very real short term costs.

As a culture we must somehow figure out how to rationale balance these issues and make rational and modest choices, because if we don't we will be making the irrational and arrogant choices and they may yet kill us all.