Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

Comment author: Viliam 05 May 2016 09:05:17AM *  1 point [-]

Worst case is IMHO that a new person will be created in an old and damaged body.

Comment author: time7 15 June 2016 12:48:51AM 0 points [-]

How would you tell, what would this "new person" theory predict differently then the old person theory.

Comment author: Lumifer 05 May 2016 02:29:45AM *  2 points [-]

The existence of that low prior is the proof that it's very likely false

I think you're trying to double-dip :-) The prior itself is a probability (or a set of probabilities). A "low prior" means that something is unlikely -- directly. It does not offer proof that it's unlikely, it just straight out states it is unlikely.

And there doesn't seem to be any reason to talk about priors, anyway. It's not like at any moment we expect a new chunk of information and will have to update our beliefs. I think it's simpler to just talk about available evidence.

As a preface let me say that I basically agree with the thrust of your arguments. I am not a Christian, afer all. However I don't consider them as anything close to a "proof" -- they look weaker to me than to you.

makes supernatural claims; that is, claims which are by definition counter to all previous observations

That is not so. Supernatural claims do not run "counter" to previous observations, they just say that certain beings/things/actions are not constrainted by laws of nature. Wright brothers' airplane was not "counter" to all previous observations of transportation devices with an engine. Recall Clarke's Third Law.

Not to mention that "all previous observations" include a lot of claims of miracles :-)

its core claims (and future predictions) are similar to many sets of (mutually contradictory) claims made by many other religions

Yep. But there is a conventional explanation for that (I do not imply that I believe it): different traditions take different views of the same underlying divinity, but find themselves in the position of the nine blind men and the elephant.

This point will also need to explain why large civilizations (e.g. China) did NOT develop anything which looks like monotheism.

the average probability of any specific branch of Christianity would still be low

That's a wrong way to look at it. Imagine that you have an underlying phenomenon which you cannot observe directly. You can only take indirect, noisy measurements. Different people take different sets of measurements, they are not the same and none of them are "true". However this does not mean that the underlying phenomenon does not exist. It only means that information available to you is indirect and noisy.

it's likely that all sects' beliefs had human causes

See above -- different people might well have human reasons to prefer this particular set of measurements or that particular set of measurements. Still does NOT mean there's nothing underlying them.

it's because many people are Christians

Well, and why is that? Why is Christianity a huge world religion? It started with a small band of persecuted Jews, why did it spread so?

Comment author: time7 15 June 2016 12:48:19AM 0 points [-]

This point will also need to explain why large civilizations (e.g. China) did NOT develop anything which looks like monotheism.

For the same reason large civilizations didn't develop anything which looks like science.

Original thread here.

Comment author: James_Miller 05 May 2016 12:32:03AM *  12 points [-]

Very little published academic literature exists on the consequences of divestment.

This is because most people who study finance would put a very high probability on the consequences being zero. If my college refuses to buy from a firm it hurts that firm a little, but if it refuses to buy stock in a firm it does that firm zero harm. The best evidence is that while firms frequently advertise to get people to buy their products, they almost never advertise to get people to buy their stock. The value of a firm's stock is determined by what the big players in the market think are the long-term fundamentals of this stock.

Comment author: time7 15 June 2016 12:46:31AM *  0 points [-]

The best evidence is that while firms frequently advertise to get people to buy their products, they almost never advertise to get people to buy their stock.

I've seen adds that I suspect were actually targeting potential investors. Granted this was followed by the company in question exploding spectacularly a few years later with the executives being charged with fraud.

Original thread here.

Comment author: time7 15 June 2016 12:45:37AM 1 point [-]

The AI might eventually conclude that the "JFK" in it's prior didn't refer to President John F. Kennedy but to some other guy with the initials JFK.

More interestingly, if the prior was specified as "The president isn't assassinated on November 22, 1963" it might conclude that John F Kennedy wasn't the president, it might look at the election of 1960, notice that it was one of the closest in American history and that there was some funny business with the election in Chicago and Texas, and thus conclude that the legitimate president was Richard Nixon who in fact wasn't assassinated on November 22, 1963.

From here it could go in any number of directions, e.g., conclude that since the president Kennedy was illegitimate then all the laws he signed as president were illegitimate and that it should therefore set about preventing them from being enforced. Or more generally develop a somewhat eccentric theory of what it means to be a legitimate elected official and act accordingly.

Note, that in the above scenario the AI doesn't have any "material" beliefs different from not having the not-assassinated prior but is still likely to act differently.

Comment author: time7 15 June 2016 12:45:09AM 0 points [-]

I always interpreted the quote as a witty way to tell someone to shut up.