Given the asteroid's relatively small size and how well watched it is by this point, i'd worry little even if its known odds of collision turned out to be nearly 100%.
One the one hand, the lead-up to arrival would offer a wonderful opportunity for testing asteroid deflection technologies live and in the flesh, and on the other hand, if these failed or weren't tried, we'd have a wonderful opportunity for examining an asteroid impact in real life, up close.
With an expected blast power of roughly 8 megatons and with all the surveillance around it, it wouldn't even be enormously difficult to prepare the impact site so that no lives are lost. This last step in the worst case of an actual unstoppable impact might not even be necessary, since most of the Earth is devoid of humans and could handle this without a hiccup..
The asteroid's orbit is not changing, so the likelihood of it hitting Earth is not changing.
What is changing is our knowledge of its orbit, which improves as new observations are added over time. Doing that shrinks the volume spanned by all trajectories compatible with the data. While the uncertainty is still large compared to the volume of the target (Earth), this shrinkage removes empty space, so it increases the odds. If the actual orbit does not intersect Earth, the odds will start going down once further shrinkage removes trajectories which do.
The odds increasing from 2% to 3% mean it's 3% likely to hit us, which is more than 2%, yes.
But the steady increase so far does not imply a continued future increase, that the true probability is much higher. He is addressing a normal human tendency to predict trends.
One the one hand, the lead-up to arrival would offer a wonderful opportunity for testing asteroid deflection technologies live and in the flesh, and on the other hand, if these failed or weren't tried, we'd have a wonderful opportunity for examining an asteroid impact in real life, up close.
With an expected blast power of roughly 8 megatons and with all the surveillance around it, it wouldn't even be enormously difficult to prepare the impact site so that no lives are lost. This last step in the worst case of an actual unstoppable impact might not even be necessary, since most of the Earth is devoid of humans and could handle this without a hiccup..
I’m struggling to understand the alleged difference between “odds increasing” and “being more likely”.
Isn’t that exactly what it means?
What is changing is our knowledge of its orbit, which improves as new observations are added over time. Doing that shrinks the volume spanned by all trajectories compatible with the data. While the uncertainty is still large compared to the volume of the target (Earth), this shrinkage removes empty space, so it increases the odds. If the actual orbit does not intersect Earth, the odds will start going down once further shrinkage removes trajectories which do.
But the steady increase so far does not imply a continued future increase, that the true probability is much higher. He is addressing a normal human tendency to predict trends.