You are probably right - the gravity needed to eject Earth completely out of its orbit must be massive. A smaller body might still be able to interfere and change Earth’s orbit into something catastrophic.
I always wondered if with enough processing power and knowledge of all the objects in the solar system you could throw a space pebble close to a slightly bigger pebble to make that pass by a bigger pebble, which in turn affects a larger pebble and so on until you are moving astroids, moons and planets.
Undoubtedly you could, but it would take basically perfect knowledge of the location and velocity of every pebble in our solar system and require about as much computing power as it would take to figure out how to wave your arms in such a way that 100 years later a hurricane strikes your enemies
Right, AI may be the reason for an economic collapse, as people rely too heavily on “machine learning”, or any other field where the computer’s results are favored over a humans.
It could even be responsible for a nuclear apocalypse, if we choose to put machines in charge of determining if we should launch missiles.
But they are not going to do this on their own, they are not going to take power from us.
I appreciate the interest, but like others on the board, I like to remain hypothetically anonymous. Hypothetically, because of your search around for model inference algorithms, then you’re bound to find my stuff. There are not that many people working on it.
Thanks. I’ve already read a few articles using that search. My apologies, it should have occurred to me that your name would be associated with the article. My bad. I guess I’m too used to Facebook.
It depends on the context and what question is being asked.
If you were to ask me, what is the A number 1 priority threat we should be focused on, sure it’s climate change.
But the OP is asking about extinction level events. To that question we have to think about far more severe (and far less likely) scenarios.
I’m talking about people who don’t make the least concession to lifestyle changes or political action for addressing climate change. So they don’t even consider it a threat in the first sense.
Climate change isn’t the only factor driving the current major loss of diversity which has been characterized as the sixth mass extinction, but it’s a major one. I would consider an existential threat to humanity not just those things that may cause our total extinction, but also those capable of driving profound and fundamental deleterious changes to life as we know it. Climate change has that potential.
In that case the answer to the OP is “none of the above”, in terms of likelihood of total extinction in any foreseeable future, and the discussion ceases to have any interest or meaningful relevance to the most important issues of our time.
It’s very difficult to wipe out all of humanity in the short term of geophysical timescales, but climate change is the most significant disruptive threat to life as we know it, and the greatest existential threat ever faced by humankind even if many of our progeny survive something like the P-E thermal maximum event and manage to eke out a degraded existence.
No worries. I don’t mind somebody asking about my work, it is kind of cool really that somebody might think my research would be worth reading.
The paper was interesting. I didn’t read all of the proofs as those need to be read slowly, and it would take longer than cooking and eating breakfast. The paper is describing an approach for converting a log of the events from software (called a trace) into a finite state machine (FSM). A FSM is simply a list of states that in which a model can be, and a list of transitions between the states along with the event (or input) that triggers the transition. The approach is simple. They simply scan the log and say that if Event X follows Event Y in the log, than a transition is added to the FSM from X to Y. There’s some voodoo (with adding the alpha and omega symbols) going on that I would need to dig into, but they state that the main difference in their approach is the resulting model is declarative instead of procedural, which makes for a simpler model (debatable).
The difference is a procedural specification is a series of steps to accomplish a task, while declarative is a high-level statement of the task. For breakfast I had a egg and bagel sandwich.
Procedural:
Turn on stove element
Place frying pan on element
Get egg from fridge
Get bagel from cupboard
Put bagel in toaster
Push down activator on toaster
… and so on.
Declarative:
Make an egg/bagel sandwich.
Sometimes it is useful to see the details of a procedural specification, but when dealing with complex systems it becomes overwhelming.