New plastic pollution discovered: 'Plasticorals' coral rubble fused with plastic

If you want specifics you might check out this paper which concluded that "runaway greenhouse could in theory be triggered by increased greenhouse forcing" (https://www.nature.com/articles/ngeo1892). It's going to happen eventually (https://en.wikipedia.org/wiki/Future_of_the_Earth#Loss_of_oc...) so all we'd really be doing is speeding up the process.

> Moving the planet and pulverising with an impact large enough to kill everything are firmly in the bounds of not currently feasible.

I mean, it sure seems unlikely, but I'm not as qualified as Dr. Greg Laughlin and his team at NASA, so without at least some kind of evidence indicating that we couldn't do it, I'm okay with trusting them on it being "not at all far-fetched".

'The technology is not at all far-fetched,' said Dr Greg Laughlin, of the Nasa Ames Research Center in California. 'It involves the same techniques that people now suggest could be used to deflect asteroids or comets heading towards Earth. We don't need raw power to move Earth, we just require delicacy of planning and manoeuvring.' (https://www.theguardian.com/environment/2001/jun/10/globalwa...)

Obviously, their plan wasn't to destroy all life on Earth, but to just move the Earth a helpful distance away and then stop, but even in that case they were concerned about sterilizing the planet.

The plan has one or two worrying aspects, however. For a start, space engineers would have to be very careful about how they directed their asteroid or comet towards Earth. The slightest miscalculation in orbit could fire it straight at Earth - with devastating consequences. It is a point acknowledged by the group. 'The collision of a 100-kilometre diameter object with the Earth at cosmic velocity would sterilise the biosphere most effectively, at least to the level of bacteria,' they state in a paper in Astrophysics and Space Science. 'The danger cannot be overemphasised.'

> We would have to be better at designing than evolution to make Grey Goo

Evolution is notoriously bad at designing. It's perfectly fine with a poor or half-assed job as long as the evolving thing typically manages to survive just long enough to make a baby or two. Survival rates can even be outright dismal and design choices absurd if the evolving thing pops out enough babies quickly.

Even a drunk engineer could have done better than evolution when designing a giant squid brain (http://ldtaxonomy.com/podcast/giant-squid/) or a hyena clitoris (https://www.wired.com/2009/07/st-best-5/). While I'm on the subject, putting the ostia at the top of the maxillary sinus instead of the bottom was a dick move.

There's a massive difference between intentionally designing something to self-replicate like crazy until all life on earth is either consumed or crushed and just settling for a replication rate that is "good enough, mostly".

GP asked for a possible scenario and I think this is a valid answer.