Are you saying that Ozempic allows you to escape society or that all the fat people are actually just druggies? Like, either way it's wild, but I'm struggling to come up with a more charitable take.
That's true, but the fact that it wasn't easy created a huge industry of fake information regarding alternative explanations of weight loss from CICO. It's nice to have some more obvious evidence to convince people that no, your body can't magically move and exercise for free when you eat less, among other strange claims I've heard from people struggling to lose weight. People stuck in that rut of not being able to eat less were prone to being deceived by countless weight loss products when the real solution (also hard mind you, but at least more viable) would have been to focus on which activities and diets actually allowed them personally to maintain a calorie deficit.
And as a bonus, now they also have one. A lot of people didn't even believe that was the solution though, and the belief change is important.
The "may" in that statement (may be moving towards...) is doing a lot of heavy lifting, but I think it's true that if you want to predict future harmful manosphere beliefs then one of your best predictors is current actions aligned with the top of the funnel. That doesn't even necessarily mean those actions are bad, but the combination of all of them is suggestive of a certain reading diet.
Separately, are they all good? Testosterone has a lot of negative side effects, and I know I certainly wouldn't mind having about half as much. Jawlines and fiber have similar issues when focused on in the wrong way. Eating healthy, feeling good, and looking good are mostly orthogonal to those trendy behaviors. If they work for you then great, but there are other options too, and I don't think she's saying that you shouldn't want to eat healthy, feel good, and look good. Indeed, she explicitly says that fibermaxxing is bad because it replaces a healthy diet. She agrees with at least some of your goals and really only (potentially) disagrees with the implementation.
Re point (1), it's definitely not just guns. In the US:
- Suppose a vehicle chooses not to stop. What's the safest way to complete the stop? High-speed chases are dangerous. The eventual spike trap or pushing the tail end of the car or whatnot is dangerous. The fleeing person won't be in a sound state of mind for safe driving even if not followed. Toss in a little human error like forgetting to turn your lights and sirens on, and these are predictably extremely dangerous. Just under 1% end in a fatality.
- There's an elevated risk of wrecks, side-swipes, and whatnot by virtue of being on the side of the road, often in dangerous areas.
- Most assaults on police officers don't involve guns. There are something like 7-8k assaults on officers due to traffic stops annually, despite knowing they carry guns. A little over a hundred of those result in serious injury to the officer, most of which don't use guns in the process.
- Don't forget injury to the passengers. 100k/yr are injured by the police at a traffic stop. Again, that's usually without guns. And why wouldn't they be? The police have nearly a carte blanche, and they're being placed into an environment where they know they're at risk, with perceived risk being higher than actual risk.
You mention the UK, so how do they fare? They have 100x fewer traffic stops, so I'll scale up the numbers accordingly. In particular, that means they have 2.5x officers injured just from vehicles being rammed during stops than the US has assaults on officers in total. That's suggestive of gun violence not being the entire story regarding traffic stop safety.
Mind you, that's not quite apples to apples. The UK has 5x fewer people and 2-3x fewer miles driven per person per year, suggesting we're missing another factor of 5-10 when trying to figure out why the US has more traffic stops, and my first guess would be that the UK police have a selection bias toward more serious offenses, so the higher safety rates of the US police are at least in part due to pulling over a higher share of less violent individuals.
Nevertheless, it's not just gun violence. Traffic stops are dangerous elsewhere too. (not to be interpreted as a comment on gun policy, rather that it's annoying to see important conversations like consequences of traffic stop safety derailed by irrelevant nationalistic swipes and jabs)
How much of the violence you list is because of warrant enforcement via 'traffic stop'?
Traffic stops should be used for traffic enforcement and treated as community policing, not as the main method of hoping to stumble on things the police should have resolved via actual police work and effort.
From here [0], extrapolating based on relative rates of traffic stops vs other police-initiated contact. An estimated 32k of those required hospitalization.
Or preferred to eat. I love teaching, and I'm good at it. If it were more valued by society then I'd do more of it. As it stands, I'm in adtech, and I teach a little piano and math on the weekends.
It sounds to me like inflammatory language around a real problem. As a rule of thumb, the people who ignore leash laws are more likely than average to have not bothered to train their dogs to be safe off-leash. I've had a few dogs bolt from their owners toward my dog and I before (normal suburb, normal locations like the dog choosing to cross a downtown road to our sidewalk from theirs), teeth bared, barking, growling, etc. In my case, I'm big, strong, and scary looking, so I just lift my dog high enough the attacker can't reach, stomp toward it, and yell something like "hey," and it'll usually back off and whimper a bit. My wife can't lift our dog easily or scare other dogs though, so our dog has been bitten a couple times when she's been out walking.
There's one person in the area who seems to have a revolving door of dangerous dogs available he likes to keep off leash. Every time the police take one away for mauling an elderly man or whatever he finds another. He lives in an RV though, and CA's relationship with the homelessness problem is complicated, so he keeps getting a slap on the wrist.
- "Assaulted," both in the legal definitions and in popular usage, isn't synonymous with "attacked," and I suspect that's where some of the strays you caught in the comment section came from. Even if correct, that's a different bar to clear.
- The dog attacking is different from the owner causing the dog to attack. The latter is a higher bar to clear, and one people are going to want more evidence for. It's also incongruous when paired with the word "assault" with some other parts of your post. In particular, "let their dogs run off-leash" doesn't evoke the same mental image as "unleash their dog and tell them to attack." Your comment has the former and has language indicating the latter. People pick up on that sort of thing. For that matter, when trying to defend you to that one commenter I had to make a guess as to which interpretation of what you wrote is the most likely way to resolve that verbal inconsistency, and I might not have gotten it right. Maybe you really do get intentionally attacked by people unleashing their dogs on you. If so, why write about free-range dogs roaming around as it pertains to the attack problem?
- Active assault is also different from self-defense and from dogs accidentally picking up non-verbal cues.
And I focused on the "dogs attacking" part of your comment in that defense, but in the context of the article, your initial point was that dogs are sensitive to nonverbal clues, your evidence was that you've been attacked, yet the experiment had what I think is stronger evidence in that they explicitly controlled for that effect. Attacks happen for many reasons, potentially including nonverbal clues. Do they happen exclusively because of a sufficiently extreme sensitivity to nonverbal clues that the experiment's control wouldn't apply? Probably not? I imagine a lot of the people who gave you flack for the way you wrote about those attacks took that into account. The structure was "<A> evidenced by <B>" when that evidentiary link is tenuous and <B> is a bold statement about something that's clearly distressed you. That comes across more as an uncontrolled rant and makes people less likely to believe <B> in its entirety, whether true or not, and whether that intellectual habit of the readers is justified or not.
Independent of any linguistic nitpicking, I'm sorry you keep getting attacked. Stay safe out there.
Unwanted touching is an assault. Jumping on someone and knocking them over is attack. I am victim of both. Both are well defined legaly.
The self defense and off leash is claimed by dog owners as an "apology". It is not something I invented. Again, I am the victim here, I do not have obligation to study excuses of some street bullies! I just want to get out of there!
> Your comment has the former and has language indicating the latter....you wrote is the most likely way to resolve that verbal inconsistency..
I think i am very consistend i hate both.
> If so, why write about free-range dogs roaming around as it pertains to the attack problem?
Because it is relevant to an attack? Dog on leash can be avoided and will not ambush person from behind!
> Attacks happen for many reasons, potentially including nonverbal clues. Do they happen exclusively because of a sufficiently extreme sensitivity to nonverbal clues that the experiment's control wouldn't apply? Probably not
Some dog wizardy, perhaps telepathy. Dog can sense owner is scared of ducks, so it hunts it. Perhaps they can pass whst they hesr to owner.
I simply do not buy anything about dogs, if owner is in the same room.
Definitions vary from state to state, but it's common for the legal definition of assault to require both willfulness and the knowledge that application of force will arise from your actions. You seem to think that bar has been reached by nonverbal cues. It might have been. It's different from just unwanted touching (indeed, often not actually requiring contact). Legal definitions aside, "assault" in popular language is also not just unwanted touching.
What your story does show is negligence causing injury. That's still bad. Nobody's condoning that.
That all said, I don't have a dog in this fight so to speak. Best of luck.
> but it's common for the legal definition of assault to require both willfulness and the knowledge that application of force will arise from your actions
Based on their other comments, they apparently do actually mean it like that: dog owners are intentionally using their dogs to attack runners, then claiming "self-defense" out of fear of runners.
1. One view of the field treats Maths as directly solving real-world problems. Under that lens, yes, solving problems is directly advantageous for the field. Even so, it's not necessarily a good thing for the participants. Suppose the incentive structures require tackling hard, long problems, the field attracts people who enjoy that sort of thing, it's hard a-priori to tell a hard problem from an aggrandized brute-force proof search, etc. Even if directly producing proofs is the desired output for the field, a capricious machine occasionally solving 40% of the big problems the field is working on is still demoralizing. On a human level, if a society has organized around producing 100yr+ problems and bestowed on you the privilege of tackling a few of them having the last 40yrs of your work solved by a fancy calculator is never going to feel good. Sure, that means that in retrospect that was a problem we could have waited to devote human effort toward, but you couldn't possibly have known that at the time. You worked hard, made locally correct decisions, and later found out that the work was worthless. Beyond that, I haven't seen any model of "Maths 2.0" produced where that issue goes away. We still want people to devote their lives to hard problems. The machine will still advance and solve a moderate fraction of them at random, killing many of the field's life's work. What motivation do you have in that environment to pursue those challenges in the first place? Why not pursue adtech or fintech and make the world a worse place instead? Solving 40% of the problems isn't +40% throughput; it's -60% as the rest of the field leaves for greener pastures, -inf when nobody's around to construct the right problems in the future..
2. "Solving problems" isn't directly the point of most of the field. That's true both from the perspective of most problems requiring many intermediate results (solving one useful thing requires solving many less useful things) and from the perspective of building the machinery capable of solving problems being much more important than the solutions themselves (a properly presented proof leveraging clever, new techniques is infinitely more valuable than whichever thing it proved -- indeed, the Millennium Prize problems were chosen precisely because it was assumed they would create such techniques, and for the human-solved ones they did -- without new techniques, eventually the field caps out at whichever frontier of knowledge we're currently stuck at).
3. A third view of the field treats it more like art. A machine producing behind your back the painting you were envisioning, or at least a close enough simile to fool the masses, doesn't advance "art" in any meaningful sense. It's extremely demoralizing when flaky cast tools are passed off for proper forged variants (your work is better, and it's sad to see people choosing a wrench they'll later snap in half instead), and when at a creative level people are accepting mediocrity instead of something objectively better.
Yeah, it's not the non-determinism; it's that they don't do what I say. I can't imagine writing a nontrivial program without, at a minimum, reading and editing the pieces as they're being produced. People aren't complaining about non-determinism; they're complaining about the results being non-deterministically sometimes wrong.
That actually would be significantly better, because then you could work out under what conditions they were wrong and correct for it. But being stochastically wrong means you have to check everything, because you never know when the errors will creep in.
If it were deterministic but chaotic, which AFAIK is the case for LLMs running at temperature 0, then you would have to likewise check every time you changed anything and couldn't meaningfully correct for it.
If the results are non-deterministic, unless I'm happy with literally any possible result, then yeah, I'm going to be unhappy with all of the other results. That's what non-determinism means. What would being right always with non-deterministic results even look like in terms of code output?
Compilers have a ton of hard optimization passes, and many of those are solved well enough via stochastic algorithms. Nobody actually fully optimizes register allocation or whatever, and a few random restarts can go a long way toward some better performance.
Every professional compiler I've ever used has given me the ability to choose what optimization level to use. No professional LLM I've ever used has given me any choice whatsoever to modify the temperature.
This argument is conflating determinism as in "byte-for-byte reproducible builds" with determinism as in "builds are semantically correct".
The fact that a traditional PL compiler output can have different performance, based on stochastic optimizations, is not in dispute. But, we consider the compiler flawed and inappropriate for use when it will randomly produce incorrect machine code. Under normal circumstances, we can infer that "correct" source code will produce "correct" builds every time it is fed through a valid compiler.
The problem people have with slop and vibe coding is that there is no formal semantics to even define correctness of the LLM/agent output.
With no formal semantics to constrain the outputs, you have to treat each prompt/inference run as a fuzzy search. What it "finds" may or may not match your intended semantics. It is merely a candidate that might correspond to your ambiguously specified semantic intent.
It is magical and lazy thinking that leads people to use this fuzzy search process in "I'm feeling lucky" mode.
Yes? And stochastic methods are one way to approximate optimal NP solutions. In the context of register allocation, you'd be most likely to see that when modeling actual runtime optimality rather than minimizing spill or whatever. Mainstream compilers don't do that, but mainstream compilers don't incorporate many SOTA techniques.
1. Programmers sometimes do something other than what they're told, so what's the harm in letting an LLM be disobedient instead?
2. Managers sometimes don't ask for the right things, so why not let an LLM figure things out?
To both points, those observations are literally the reason why we have software engineers writing code instead of management. The job of fully understanding the implications of all requirement interactions in excruciating detail is not a management responsibility. In the context of using LLMs as a higher-level compiler, I don't think people are arguing that management should be the people writing that higher-level code.
Separately, the failure modes are very different. If I ask a SOTA agent and a SOTA harness to construct a minimal sandbox for executing subprocesses without network access using OS sandboxing primitives to ensure you don't have Scun*horpe-esque ways of sneaking networked commands through (and assume I expanded that into a proper prompt instead of the high-level summary I'm using to convey the idea here), there's a nontrivial chance it implements some of the sandbox by string matching the commands instead of using proper OS primitives. It'll have thousands of tests showing it works, and it'll still be garbage even just from a requirements point of view and ignoring code quality or everything else we're being told to ignore when using them as higher-level compilers. There's also a nontrivial chance that if you ask either the same or a new session to check on that property it will conclude that the sandbox was implemented correctly. Good human engineers will either get the job done or tell you that they can't for some reason or another. That "looks like it succeeded but actually failed hard and you can't tell unless you read the code" error case is new to LLMs. Yeah, we write bugs from time to time (as do LLMs), but complete failures to understand the core product so badly that fixing the "bug" requires a full rewrite, one likely to introduce its own similarly catastrophic rewrite-inducing bugs, are a different beast altogether.
And again, I use LLMs a lot, almost exclusively, but they have limitations, and failing to recognize those isn't conducive to success. The frontier keeps moving, so you might find these current observations irrelevant eventually. People are choosing to not read the code today though, and that's bad.
>In the context of using LLMs as a higher-level compiler, I don't think people are arguing that management should be the people writing that higher-level code.
>1. Programmers sometimes do something other than what they're told, so what's the harm in letting an LLM be disobedient instead?
1. People spend a lot of time talking about how LLMs often do something other than what they are told, as if people forget that they aren't especially good at doing what they're told either.
[0] https://maps.app.goo.gl/xnJUmJArCeaDmKem9
reply