“Take a bath, please repeat my suggestion back to me.”
Encouraging people to think thoughts that inspire them to run screaming through the streets naked is a value I wish to promote.
“Take a bath, please repeat my suggestion back to me.”
Encouraging people to think thoughts that inspire them to run screaming through the streets naked is a value I wish to promote.
Saint Peter of Verona, patron saint of inquisitors, practiced this method when dealing with suspected heretics. By allowing himself to have a crisis of faith when confronted with the sincerity of his opposition, his beliefs came out stronger in the end and we’re often persuasive. Saint Peter not only never lost his faith, but through his example, inspired his assassin to devote his life to the Church.
I suggest instead finding an unforgivable sin within the religion you are seeking to escape. Then committing that sin gives you a personal incentive to build a belief structure that does not require good standing within that religion. For Christianity, simply saying ‘I deny the holy Spirit’ can be sufficient to meet this condition. For Islam, saying the words, Allah is not god, and Mohammed is not a prophet might work, but I’m less familiar with Muslim theology.
Is there such a cardinal sin in ‘rationality’?
What evil can be perpetrated by AGI that cannot be perpetrated by a sufficiently capable human or group of colluding humans?
Leo Szilard could probably have built a bomb that would wipe out the human race, we are still here, and do not credit that to the success of developing a ‘Friendly Hungarian’ or the success of the ‘Hungarian Safety’ research community. Arguably, Edward Teller was a ‘slightly unfriendly’ Hungarian, and we did OK with him too.
He produced a then novel scenario for a technological development which could potentially have that consequence: https://en.m.wikipedia.org/wiki/Cobalt_bomb
He also worked in the field of nuclear weapons development, and may have had access to the necessary material, equipment, and personnel required to construct such a device, or modify an existing device intended for use in a nuclear test.
I assert that my use of ‘sufficiently’ in this context is appropriate, the intellectual threshold for humanity-destroying action is fairly low, and certainly within the capacity of many humans today.
Addressing your question, Szilard’s political action: https://en.m.wikipedia.org/wiki/Einstein–Szilárd_letter directly led to the construction of the a-bomb and the nuclear arms race. The jury is still out on whether that wipes out the human race.
I assert that at present, the number of AGIs capable of doing as much damage as the two human figures you named is zero. I further assert that the number of humans capable of doing tremendous damage to the earth or the human race is likely to increase, not decrease.
I assert that the risk posed of AGI acting without human influence destroying the human race will never exceed the risk of humans, making use of technology (including AGI), destroying the human race through malice or incompetence.
Therefore, I assert that your If-Then statement is more likely to become true in the future than the opposite (if no humans have the capability to kill all humans then long-term ai safety is probably a good priority).
Bifur should tell everyone that he is going to try to wake the Balrog, and dig directly towards it, openly advertising his intent to be the first to wake it. Spreading rumors that he intends to yoke the Balrog to a plow, and that he alone has a specific plan for accomplishing this would be helpful too.
The action taken to control the insanity of that one crazy dwarf might prevent the catastrophe outright.
Thank you! So, the path of purposeful self-deception is not the road to higher rationality, no matter how well it happens to work.
To use the monkey riding on the tiger analogy for human cognition, I wonder which is more effective. The monkey putting the tiger in a pen and swinging through the trees alone...Or the monkey that ties a steak to a stick and rides the tiger.
You could implement this today. Found a leadership PAC called ‘Smarter Government’, use big data to identify the set of people who are the correct age and meet residency requirements for the office the PAC is pursuing, send them all phone calls asking if they’re interested in elected office, administer the battery of tests, provide the lucky winner with 1-200k seed money from PAC donors and a competent staff.
I think the testing measures would have to include stuff other than straight IQ, because ultimately, this person has to win an election. Probably by any available means including deception and emotional appeal.
Unfortunately, that criteria is probably impossible to do away with completely. Much of the job of a politician requires positive action on the part of the office holder. ‘Does not want the job’ or ‘unwilling to perform the task if chosen’ is probably a disqualifying factor, so may as well select against it early in the process. Perhaps a question emphasizing community service (are you interested in taking on a significant responsibility in your community?) would work, but ultimately, it would be impossible to keep the screening process completely secret.
Oog and Og are sitting in the forest. Oog says ‘man someone could build a fire and totally destroy this whole place, we need to devote a lot more energy to stove research before that happens.‘. Og says ‘sure, fire, OK, that’s sci-fi stuff, I’m going to go gather some berries while you waste your time with heat flow calculations and the stove safety problem’.
Oog doesn’t like being brushed off, so he decides to try a different tactic. Og returns to see Oog waving a brightly burning torch around in the air, dangerously close to the dry leaves and twigs of their shelter. Og’s reaction is far less skeptical than before: ‘Oog! You will kill us all, stop waving that thing around until we have at least a pit dug!’
So yes, the best thing you can do to popularize the cause of AI safety could be building an obviously unsafe AI and doing something demonstrably dangerous with it.
Taking the analogy further to a community of tiger riding monkeys...The monkey that waves the steak on a stick in front of some other monkey’s tiger probably has a future in marketing.
The monkeys who decide to pen their tigers may have a problem, the tigers are still present, may be unhappy about their confinement, and after a time, the monkeys may not watch them as closely as they should...
As a case in point, I give you the prevalence of polyamory in the rationalist community. Historically, polygyny has been a feature of insular communities that wanted to become more insular. Is polyamory serving its purpose as a strong social barrier to entry for the high table of the rationalist community, or is it really just pure rationality at work?
This problem is an easy one to solve. Implicit association tests are effective at discriminating pedophiles from non-pedophiles.
Have ‘age of consent’ set as a community norm (keep it legal folks!) and ‘must score under community-assessed maximum tolerable score on the pedo-IAT’ as a condition of moving in.
You’d be the first neighborhood association to take this sort of strong, invasive(?) measure for preventing child abuse in a close knit community, I don’t think they do this in Celebration, Florida (or whatever the Disney World community is called)
https://www.ncbi.nlm.nih.gov/pubmed/23613137
Note, any implementation should also include eye tracking, or another analytic to detect a user looking away and clicking at a constant rate. Looking away and clicking to advance at random or at a constant rate is the only mechanism for defeating an IAT.
How do I apply for a slot in the house? I’m tired of living by the sword and would love to relocate to a tech hub but don’t know anyone.
Your corporate plan would likely work. White nationalist Craig Cobb attempted to purchase large tracts of land in Leith, ND with the express purpose of providing them exclusively to white nationalists. Some aspect of this plan appeared to get him around the Fair Housing Act.
I believe he was run out of town along with his little club, so the best advice would be basically ‘avoid advertising outside of rationalist circles’, and don’t antagonize your non rationalist neighbors.
To post title: Yes. See this discussion of quantum interference (decoherence) in human decision making: http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0168045
Utility function +/- 25% against most uncertain prospect, in favor of prospect directly opposite most uncertain prospect. Add an additional +/- >5% as more information becomes available.
Somebody use that 25% in a back-prop algorithm already plz.
The IAT is part of a somewhat widely used assessment of sex offenders, the Able Assessment, and is less invasive than penile plethysmography (ref1). Unfortunately, IATs have been shown to be unhelpful for identifying female child sex offenders, as their cognitive approach to offending is different from that of men (‘I was coerced by a man/lonely and horny’ vs ‘entitled and attracted to the bodies of children’) (ref2).
There is likely a false positive rate for an IAT, enough that it is relegated to the realm of polygraphy, and inadmissible in court...but I am not particularly concerned, as it is likely not large enough to render the test worse than random, and for a community like this, given that no additional discrimination will be applied beyond ‘please live somewhere else’, males in this specific, vulnerable community should be fine with submitting to an IAT. Given the ‘male coercion’ factor in female sex offenders, denying access to men who ‘fail’ the IAT would probably reduce the liklihood of female offending as well.
The reproducibility crisis is real, and most psychometric tests are lousy for a number of reasons, but it is possible to extract data that is useful, though not perfect, for making decisions. This is not a fire-and-forget solution to the problem, but in concert with normal behavior intended to reduce harm, it will hopefully help prevent the ‘Rationalist Baugruppe’ from devolving into a ‘Rationalist Pitcairn Island’
Assessment survey: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2993520/ Survey on women: https://beta.bps.org.uk/news-and-policy/developing-assessment-and-treatment-practices-female-sexual-offenders
For the sake of argument, some numbers to match the assumptions you named. Let’s base these assumptions on some numbers available to Americans today, rounded to even numbers in the direction least favorable to my argument.
Percentage of population that are psychopaths: 1% (two orders of magnitude more non psychopaths than psychopaths exist today) Probability of being victim of violent crime varies a lot based on demographics, 10 per 1000 per year is reasonable...so 1% Power consumption of human mind: 20W (based on the human brain, we will not hit this immediately, but it is a design goal, and may even be exceeded in efficiency as we get better) Power consumed by typical American household: 900kWh per month (100 years in brain-seconds) Number of humans available for uploading: 10 billion.
Over a hundred thousand years, that’s a lot of terrible people, a lot of spare capacity for evil, and a high probability of everyone eventually experiencing a violent crime, like upload-torment. Changes to those numbers unfavorable to this scenario require incredible optimism about social developments, and pessimism about technical developments.
I feel like just about anyone, even without a stanford prison experiment like environment, can muster up the will to leave a lightbulb on for a while out of spite.
Arguably, once ‘captured’, the aggregate total time spent experiencing torture for a given future copy of you may vastly exceed the time spent on anything else.
Anyone who argues in favor of ‘merciful’ euthanasia for people on the way to horrific medical problems would likely argue in favor of secure deletion to avoid an eternity in hell.
As stated in a separate comment, the human mind runs at 20W, so that’s probably a reasonable design goal for the power consumption of an emulation. Keeping a few copies of minds around for torture will eventually be a cheap luxury, comparable to leaving a lightbulb on.
I discover evidence that some sadistic jerk has stolen copies of both our minds, uploaded them to a toture simulation, and placed the torture simulation on a satellite orbiting the sun with no external communication inputs and a command to run for as long as possible at maximum speed. Rescue via spaceship is challenging and would involve tremendous resources that we do not have available to us.
I have a laser I can use to destroy the satellite, but a limited window in which to do it (would have to wait for orbits to realign to shoot again).
Would you be upset if I took the shot without consulting you?
What are dangerous things that a malign AI superintelligence can do, which a large enough group of humans with sufficient motivation cannot do? All the “horrible threats” listed are things that are well within the ability of large organizations that exist today. So why would an “AI superintelligence” able to execute those actions on its’ own, or at the direction of its’ human masters be more of a problem than the status quo?