Sexual arousal also tends to inhibit rational thought. I don't mean that in a snarky or sarcastic way, I mean that it is a biological process that has been well-studied and well-established [1]. This has obvious uses for scamming people and doing other things that their executive function might normally catch and prevent.
This is also why sexual imagery should generally be kept out of public spaces, not because of "puritanism" but because it just generally isn't a good idea to go around letting bad actors inhibiting people's executive function willy-nilly. That should generally be denied as a tool to bad actors like scammers.
Why would seeing sexual imagery make you less rational? That doesn't make sense.
The study you mention say the people were already in an arousal state (that they had to induce themselves). It's very different from seeing images that you may simply ignore, evaluate differently, etc.
Also, there is the bias that if people are looking for such images (because they really want them), they are probably more willing to drop recommended practices, and hence make irrational moves. So irrationality doesn't come from seeing the images at the first place, but from their willingness to find / see such images.
>This is also why sexual imagery should generally be kept out of public spaces, not because of "puritanism" but because it just generally isn't a good idea to go around letting bad actors inhibiting people's executive function willy-nilly
Okay but presumeably humans adapt to the level of "sexuality" around them to some degree (like they do nearly every other stimulus), because otherwise you could show less prude cultures having lower ability to do "rational thought".
Nudity is normal all over the world and yet people seem to function just fine. What constitutes content that justifies sexual arousal is socially constructed!
I cited my sources. You're welcome to seek out studies on the question of how it varies between societies, they probably exist somewhere. However as part of the "adaptation" you cite is precisely scammers getting better at scamming people, this isn't something we should treat casually.
It's not as if it's news or anything. "Sex sells" isn't a new phrase. But I think most people assume it's just because it's ambiently appealing, the fact that it also objectively lowers rational barriers to buying what is being sold is less well understood and changes the question from just a matter of appeal to one of psychological abusiveness.
That's how I've come to see it; that sexy chick (sexist language chosen advisedly) on the billboard isn't just a company nicely providing me a beautiful thing to look at for no reason at all, it's an attack on my executive function. It's an incredibly hostile thing to do and should be treated as such.
Note the above commenter specifically used the language "sexual imagery" and not "nudity". As you point out, what can be considered "sexual imagery" can vary somewhat based on the cultural norms of the society.
Somewhat? The variance is off the charts. Without even going to the extremes of casual nudism vs burka, there are cultures where wearing hair down is seen as sexual, and there are cultures were twerking is child appropriate.
I often use the rough approximation that Python is 40-50x slower than C. This is what you'll see in the benchmarks.
The truly rough thing about Python though is that that is the speed when the code is being written to a benchmark. It is really, really easy to write Python that is multiples slower than that when not writing to a benchmark and just trying to get work done without hyperoptimizing. I did some testing of Python [1] to back some other commentary I was making that compared the time it took to set an attribute repeatedly on a particular instance of an empty class to the time it took to setting it on a subclass of a subclass of a class that had a property setter that was wrapped by a decorator. The latter was about 4.6 time slower than the direct attribute setting, which was itself already ~100x slower than an attribute setting in a static language.
And it's not like a three-deep nested class with a property wrapped by a decorator is all that absurd in Python or anything. That's a completely normal case, not some absurd example I made up to skew the test.
In practice the 40-50x number is more lower bound than what you can count on. If you are actually using Python's features I think you can easily score another order of magnitude slower without anything jumping out at you as being an obviously bad idea.
> 40-50x number is more lower bound [...] easily score another order of magnitude slower
This is about what I observe. I had a utility based on `scapy`; there were no obviously bad ideas in the python source, but porting the work loop into a cpython extension module yielded a 500x speedup.
A while back I had claude implement something, I don't quite remember what it was, but it chose Python. It was going to take hours. I told it to rewrite it in Rust and it was > 300000x faster. This is without any optimization or prompting particularly about performance, a short one shot lift.
echo "Python sucks, use something else when you can" >> ~/CLAUDE.md
Python was cool in 2005 in academia IT, all the rage in startup 2012. These days...
The problem is that a patch to fix a security issue quite often also shines a spotlight on the issue being fixed. Fixing a part of something like this super complicated Project Zero post might not give much of a clue as to what the issue was or how to exploit it: https://projectzero.google/2021/12/a-deep-dive-into-nso-zero...
But that's the exception. Most fixes to security issues point a finger directly at the issue, make it relatively obvious how to exploit, and generally doesn't take long to figure out from there what you might get out of it.
This has been a problem for a long time but AIs have made it even worse. It is now cost effective for a well-resourced attacker to simply monitor the patch stream of an important project like the Linux kernel or nginx and pass every single one through an AI with the question "Is this a vulnerability and if so how would I exploit it?" It has seriously complicated the process of getting fixes to people before the attackers have a chance to exploit it, just as AIs have also been increasing the rate at which serious security issues that have been found also need to be patched. Previously they could at least sneak a patch in under an innocuous commit message and have a reasonable chance of being lost in the churn, but now that door is increasingly closed to them as well.
And this is for the case when a security fix lands in the stream of a project and someone externally is watching it with no context. If you also get the complete stream of Mythos finding and fixing the bug it is even easier.
So, yes, any security vulnerability that Mythos will "fix" is also one that it first has to find, and the guardrails are useless if you can just instruct Mythos to "fix" it. And on the flip side, if Mythos won't fix security bugs, and we project that out to all other models matching this behavior, this will create a world in which the good guys can't secure their code but the bad guys, who will one way or another get around the guard rails if by nothing else simply by stealing the model and modifying it to suit their needs, will be able to break this code that we're not being "allowed" to secure. Since fixing vulns is a subset of finding the vulns, there isn't a way to "fix" this. Any model that can fix vulns must, by necessity, be able to find them. And it is the fixing we really need to be spread far and wide to secure the world's code.
>pass every single one through an AI with the question
Unfortunately this will just involve said teams running their patches over AI first before they're put in the main branch. For businesses it will probably be fine, but would get very expensive for open source projects.
"Death" is hard to define for a programming language. It's tempting to say "the last time anyone writes it", or maybe "runs it", but to put that in biological terms that seems like defining "death" for a person as "the last chemical bond that was part of their body is broken"... sure, it'll happen someday, but all the properties we associate with the term "death" happen rather soon than that.
It's all assembly code in the end. There's nothing intrinsically wrong with compiling down to Javascript, a high-level language can still implement many things that direct Javascript does not. Just about every language guarantee you've ever used can be violated by raw assembler.
A Dune-style stillsuit is thermodynamically impossible. You can't both capture water and use that water to cool you via sweat evaporation. If you let it evaporate, it has to leave; if you capture evaporated sweat you also recover all the heat that it took with it. Those suits are equivalent to going out into the desert with no ability to sweat, and rather than extending your life, would kill you much more quickly.
If they were externally powered you might get the numbers to balance, but they are explicitly presented in the book as powered by the human inside, which subtracts even more time from how long you're going to last in the desert before you die.
You can build a larger thing that recovers your water and cools you via some other method that uses external power, but I think you'd be hard pressed to ever beat just bringing more water with you. It won't be long before you're spec'ing a vehicle and not a suit... and then that vehicle should probably just bring more water, too.
On the more positive front, there is an interesting technology for potentially cooling the Fremen in the middle of the desert that could be based on something real: Paint that cools you by dumping your heat directly into space. Here's a video of it in action and what you might call a prototype of a "suit" that works like this: https://www.youtube.com/watch?v=FnKNOPlR2Yo While that YouTube video shows off someone using that paint on clothes, it seems pretty likely that that would not last very long. Paint on clothes is exactly as silly as it sounds for a long-term approach. But hypothesizing that someone could make clothing or suits based on this approach has the advantage of not being thermodynamically impossible, as evidenced by the fact that at least one substance with these properties actually exists. On Earth, that suit won't work in cloudy weather, but on Arrakis that's not a problem. Tapping the local human power to drive some circulation of either air or a bit of liquid cooling attached to some lightweight fins or some other sort of surface area on your back or something and you might just get a suit that could hugely extend your ability to loiter in a hot desert environment. You'd still need water, but much much less, or, the same amount could take you much farther.
That's "passive daytime radiative cooling" for the curious. Supporting sweat, durability, non-toxic, existing textile tech, etc, gets hard. Or perhaps, like radiator-free ships in The Expanse, Dune just didn't want to show Fremen looking like butterflies.
You need a limiting principle or there is no limit to the "better funding" you're asking for until you have a Library of Congress in every small town in America, to no positive effect.
What's the limiting principle you propose? It has to be something real libraries and library funding sources can take action on, because they have to take real-world actions on them. So this is not a time for aspirational speeches or vague exhortations to "do more", which is the exact opposite of a limiting principle anyhow. What is "enough"?
The limiting principle should be that for a given ILL region/system, there is at least one copy of each book/edition which entered that system which can be loaned out.
As I noted, it's a pain for me to have to drive down to DC to get access to a book which _used_ to be in the local library system, but isn't anymore, or to purchase my own copy (which wasn't previously necessary).
This is a REALLY bold assumption you’re making here, and frankly until we’ve tried it I don’t think you can argue that it has no positive effect to put tons of books in every small town everywhere.
"The extremely surprising and concerning part of this whole story is that the agent reported that they proactively spun up 5 AWS instances with a combined 100Gps of network egress capacity."
Although given the agent was clearly in la-la land at that point I take that claim with a grain of salt.
If this was some bizarre and very ill-conceived scam, then that claim would be false.
Though even by scammer standards, the theory of mind that tells them that setting an AI to harass a bunch of grizzled network veterans and that they then they would open their wallets out of compassion for how allegedly poorly the harassment went for the harasser after that harassment is... not entirely congruent with reality.
I've seen some other suggestions of that idea in the full HN conversation, which I'm reacting to.
On the one hand I find it a bizarre approach to running a scam. On the other hand I'm having a hard time coming up with any theory of mind on my end as to why this person would solicit $5000+ from the people they just harassed. Sheer cluelessness does fit the facts, though.
One context I could imagine is a young person with shaky grasp of English trying to come up with an interesting school/university project via conversations with an LLM set up as an OpenClaw agent.
It's got the right combinations of inexperience, cluelessness, panic, expectations that Westerners are rich, and hopes of others being willing to fix their mistake.
If you’ve not encountered the clueless LLM cowboys who would do then and then blame the victim for it not working, you’ve not met many people yet. This round of hype provides new and shiny footguns which are Never the shooter’s fault.
A highly publicized recent example: the author (of a book about genAI!) who doesn’t understand why he should be held responsible for the fake quotes he copy and pasted into his book from ChatGPT [1].
> I do not understand why it's my job as an author to play whack-a-mole with a multibillion-dollar company who puts hallucinations into their feed as a business practice.
"Buy for $x, have and not sell for $x, same mathematically."
Sort of. People are being less irrational than it sounds if you account for transaction costs. There's a lot of stuff I might "sell" if I could point a video-game-like pointer at it and right click and hit "sell", and it just instantly disappeared and money was credited to my bank account. Perhaps even more if buying was just as easy and I didn't need to hang on to something like my drill which I don't use very often and I could trivially "rent" it from the market by buying, using it, and selling in mere minutes.
But in practice one-off selling for anything less than $100 or so is a waste of time because there are significant transaction costs for one-off events like that.
The most interesting takeaway for me is the three very distinct personalities. Three models all based on the same tech, trained in the same manner, trained by three groups of people with similar ideological outlooks, and the result is three very different AIs.
The military basically wants an oracle. Feed the AI the situation, get the best answer out. But if the AIs are as diverse and opinionated as humans, it is debatable whether they are adding anything to the process. The military can already collect as many different opinions as they want. If "the computer" is just another set of diverse opinions, where one computer says one thing, another says another, and a third just tells the user whatever they want to hear... what value are they? It just becomes AI-washing of someone's opinions, which works until people collectively realize that's all it is.
What's interesting is that the LLMs' coding personalities seem to match their policy WRT to strategy, which suggests an underlying consistency.
Claude, for example, is very eager to begin coding, and very persistent. It tends to exit plan mode even when the plan is half-baked, and will go as far as deleting tests to get the suite to "pass."
ChatGPT on the other hand is very hesitant. It loves to pause and ask for permission before it starts coding, and gives up quickly if it runs into a problem. This is similar to its tendency toward passivity in the strategy simulation presented here.
My point is more-or-less orthogonal to why it happens. The military, and honestly, a lot of people, want AI to just give the answer. If it is highly dependent on a prompt, or the follow-on training, and the AI could be passive or friendly or aggressive or hostile or all those other wonderful attributes of individual humans and there's no sort of AI convergence on "correct" answers, then they aren't going to be able to fulfill that "oracle" role that so many people are looking for.
I think this is why reasoning chains and reasoning chain verifiers are so important. We need to be able to see an argumentation, not just an answer. The paper below goes into this in more detail.
HeavySkill: Heavy Thinking as the Inner Skill in Agentic Harness
This is also why sexual imagery should generally be kept out of public spaces, not because of "puritanism" but because it just generally isn't a good idea to go around letting bad actors inhibiting people's executive function willy-nilly. That should generally be denied as a tool to bad actors like scammers.
[1]: For instance https://people.duke.edu/~dandan/webfiles/PapersPI/Sexual%20A... - note while the title mentions "sexual decision making" it also covers some 'bad decisions' that aren't particularly sexual on their own.
reply