It’s been said that “Nigerian prince” email scammers intentionally use poor spelling and grammar in order to narrow their funnel at the top by quickly weeding out observant and wary people who are unlikely to fall for it.
By that token, LM-generated content which looks good at a glance but doesn’t hold up to scrutiny seems ideal for scamming. I’d speculate that in the scam content generation workflow, not only is there little penalty to skipping the verification step, but since you intend to push that step onto the target and hope the result is a false positive, subtly wrong hallucination might be not only tolerable but in fact better for its purpose than what a human could produce.
I really wanted to find an answer for your challenge but it seems there aren't any answers using the regular 1-26 gematria and numbers written in their ordinary English form (without "and"). However, I did find some answers using a "zero-indexed" gematria ranging from 0-25, and with this other English gematria described here on Wikipedia: https://en.wikipedia.org/wiki/English_Qabalah#R._Leo_Gillis'...
Python 3:
>>> num_strings = ['Zero', 'One', 'Two', ... 'Seven Hundred Thirty Two' ... ] # I tried 0 to 1000. num_strings generation script sold separately
>>> def find_gematria_matches(values): return [i for i,ns in enumerate(num_strings) if i == sum(values[ord(c)-65] for c in ns.upper() if c.isalpha())]
There have existed hand sanitizer formulations containing Triclosan or BZK instead of alcohol, although those compounds are more commonly used in anti-bacterial soaps. Triclosan was particularly controversial because of its potential to cause antibiotic resistance and also for being an endocrine disruptor. During the second half of the last decade, its usage was widely restricted by regulatory agencies and it was phased out of a ton of consumer products.
The whole paragraph except for the first and last sentences may have been composed using an LLM.
> All attorneys appearing before the Court must file on the docket a certificate attesting either that no portion of [...] or that any language drafted by generative artificial intelligence was checked for accuracy, using print reporters or traditional legal databases, by a human being [...] held responsible under Rule 11 for the contents of any filing that he or she signs and submits to the Court, regardless of whether generative artificial intelligence drafted any portion of that filing
Asking the attorneys to re-acknowledge that documents they file are their official entries into the record, no matter what programs are used to generate them, makes sense in principle as a way to preempt the "algorithm told us to" argument.
Well, that's N=1. But we have seen that it's sometimes possible to bypass that kind of filter with clever prompt engineering. And because these things are black boxes, it doesn't seem possible to rigorously prove "unjailbreakability"
A couple of years ago, I was attacked by a Kiwi bot near a UC campus. This is my story.
The bot and I were moving towards each other on a sidewalk, and when I came close it stopped, as they do when sensing an object in front of them. But there was an awkward moment as I tried to go around it and it repeatedly jerked forward an inch as its motor kicked on and off. Maybe I was walking around the very edge of its radius. In any case, my behavior must have triggered some pathfinding bug, because it turned and drove right into my legs, after which it stopped and sat stationary. Luckily they're small and move slowly so it wasn't a big deal, but that memory stuck with me. Articles about Tesla pathfinding issues always bring it back to the surface.
Comedian: 45,
Political Commentator/Media Personality: 8,
Brian Redban: 5,
Health/Fitness: 2,
Scientist: 2,
Author: 2,
Musician: 1,
Pornstar: 1,
(MMA) Fight Companion: 1,
Giorgio Tsoukalos,
Kevin Smith,
Cliffy B & Johnny Cristo (can't even figure out who these last two are)
Here's what I'm wondering, and pardon me if this is a stupid question:
Why go through all of the trouble generating true randomness with a Geiger counter just to use the result as the seed to a PRNG, rather than using your true random method to generate the lottery numbers directly?
That logic assumes that the amount of music produced is constant over time, which it isn't necessarily. The population grows and the barrier to entry keeps getting lower.
My 2022 New Year's Resolution is to try out complete sobriety (caffeine excluded; this effectively means alcohol and cannabis). I've never considered myself to have a 'problem' with substances like some people I have known, but I sure have spent plenty of time in my life intoxicated alone. The resolution isn't an ambitious thing because I've been going this way for a while anyway.
Over the years I've slowly come to a realization: These substances have various effects, but at the heart what they really do is make me less aware. Sometimes I guess it's a good thing. Alcohol makes me less aware of the part of me that is self-conscious in social situations, and of how others perceive me. Cannabis makes you feel more aware of experiences, but it proves to be an illusion. I guess they're really not that bad on the balance, but as I grow older and I have spent more and more time thinking about cultivating awareness - of the present moment; of my body and mind and senses; of things in life that are truly important, and which maybe even make me anxious to contemplate - I find that I simply don't enjoy intoxication as much anymore because there's something I enjoy more about awareness.
More and more I hear this nagging voice when intoxicated. It says: "I'm bored; I'm nervous; I'm scared; I'm sad; I'm worried; I'm self-conscious; I'm restless; Someday I will die. What I'm doing right now is trying to be less aware of these things. But maybe they aren't just to be ignored or avoided. Maybe they're an adaptive part of the human mind. Maybe there really is something worth being anxious about."
Isn't CTCI basically just a workbook? It has ~20 pages of intro and then problems and solutions. So it's not really a book you 'read' anyway, it's one you work through. I think the way you get really good at data structure/algorithms problems is by doing a lot of them and there's no way around that.
I have a hard time conceiving of any reason that North and South would be different in this regard, but maybe it's possible if there is a tendency for rivers to flow towards the equator for some reason, since most of the world's landmass happens to be in the northern hemisphere.
Why does it make a difference whether your wealth is denoted by USD, Bitcoin, milk caps, or Rai Stones? If tanks roll over your capital city and your countrymen are made to kneel in a row, what good is it that you can refuse to give up the private keys? And how does the medium of exchange have any bearing on nations going to war over control of oil, rare earth minerals, strategic islands?
It seems to me you're implying that the colossal environmental damage inherent to a Bitcoin-driven world economy would be offset by the abolition of militaries, police, and banks, but I don't see how you make that jump. $trillions of military spending are used to secure resources and strategic assets and ensure national security, not to defend against bank robberies. A world without militaries or police is unimaginably different from our world in so many more ways than just the medium of exchange.
And if you can't point guns at people to make them give up part of their pile, how will you implement the CO2 tax?
I would guess consuming the amount of wood required to implement such a transaction history might end up proving to have disastrous results on the environment.
I understand your perspective. I love cooking as well and I like to tweak recipes and try to never make a dish exactly the same way twice.
There is a happy medium in the amount of information contained in a recipe. Certainly a recipe that says to cook 7 ounces of meat for 12 minutes so that it reaches 167F internal temperature is ridiculous and if that recipe will fail if not executed perfectly, then that is indeed stifling to my creativity. On the other hand, if a recipe doesn't tell me whether it means a tablespoon or a cup of flour, that actually makes it more difficult to be creative since it is unclear what I would be deviating from. I've never felt stymied by a recipe that says to use 1 tbsp of butter/flour for a roux - if I decide I want to add more then I will.
I guess my complaint is that the writer has removed information from the recipe such that it now contains less information than that generally accepted medium. A certain balance of information has become convention in cookbooks and I think it's for good reason. Removing that information to me smacks of a mindset that we programmers can be notorious for - something like "this other domain can't be that hard, I can improve it by removing all this extraneous complexity that seems to me to have no purpose." It brings to mind the parable of Chesterton's fence.
Thank you for your thoughts. I hope this makes the spirit of my previous comment more clear.
By that token, LM-generated content which looks good at a glance but doesn’t hold up to scrutiny seems ideal for scamming. I’d speculate that in the scam content generation workflow, not only is there little penalty to skipping the verification step, but since you intend to push that step onto the target and hope the result is a false positive, subtly wrong hallucination might be not only tolerable but in fact better for its purpose than what a human could produce.