> Seems like it's more beneficial to acknowledge the bad stuff than to encourage positive thinking.
Psychologists are sure it is. You should get all your traumatic experience and deal with it. You'd better learn how to remember these things without panic attacks or whatever. And the methods they use is replaying the traumatic memories multiple times while controlling the emotional state. The controlled emotional state sticks to the memories and replaces the one that was remembered before.
Well, that's the theory at least. I tried it and it kinda work but not perfectly, it may require some recurrent sessions over time if the effect fades. Though if you practice it a lot, it becomes a habit, an automatic response to traumatic memories, so any memory replay reduces the strength of the memory.
It is like a positive thinking (you get your negative reaction to the memories and replace it with a positive one... well, maybe just less negative), but it is definitely not burying unprocessed memories deep inside your mind.
> the children who drew pictures of the injuries and catastrophe, later showed fewer symptoms of stress and anxiety
I believe it is easier for kids, they are more focused on "here and now", and just replaying a memory in a safe environment has much stronger therapeutic effect than for adults. It is easier for adults to ignore the present safety and to dive deep into their past memories with all the associated emotions, so replaying memories can easily make them worse by intensifying remembered emotions.
Adults have crystallized worldviews, which were probably shaped by their traumatic memories, and it shapes their automatic emotional response, and makes matters worse, harder to change. Children are more fluid, they have more plasticity.
> The alternative they present is arguably less secure because the function pointer will remain writable for the life of the process
The article mentions this, and also points to mprotect which you can use to protect the pointer.
Why people jump to criticize without reading first? BTW, you can ask an LLM to check your critique, before posting, if you don't want to read the text.
You can always populate .init_array section with a hidden constructor. It would work just like ifunc and would always execute at shared library load time.
Yes, liblzma could have used multiple routes to take over sshd. Once you're running inside the process it's game over. The exact details, like how they used ifunc and an audit hook, are very interesting, but ultimately not that important.
I broke my leg recently. Shortly after that I've lost my consciousness. It was very painful, the body reacted with a lot of adrenaline, and after a several minutes when adrenaline was drained away my consciousness was drained too.
I experienced something like this several times, though not to the point of fainting. But this time was special in other way too: I had friends near me, they observed me through all the process and we could compare our observations later. It seems, that my memory stopped recording before I fainted. I was still operating to some extent, but I couldn't remember a thing. When asked something I grunted in answer. When one of my friends insisted that I stand up and come to a better place to sit down, I actually stand up and did several steps before stopping and slowly (and carefully) sank to the ground. (An interesting observation, my controls over my body were weakening, but I was still using them for what they worth. It fits with all other similar experiences: the limbs and all the muscles seem to be losing their strength, and it takes a lot of will to make them work.)
On the light of this, I'm very interested what proponents of the idea, that feelings need consciousness to work, would say about my half-unconscious state. Did I feel myself extremely bad at the time? Or maybe I didn't feel anything? My friends are sure that the former statement is true, but they may be mistaken by my outside looks. I personally don't remember. Up to some point I remember that I felt really bad, but the next thing I remember I look at the sky and I'm surprised by what I see (I was not in a place I expected to be). And at that moment I was pretty ok already, no more adrenaline issues, just my leg was aching.
Was I experiencing qualia is another interesting question. I'm pretty sure I was, but I'd like to hear an argument for the opposite.
Maybe it was a different part of your nervous system experiencing them, akin to a BIOS versus the operating system. The brain is a very complex and fractal thing, it is entirely possible that a more basal part of "you" took over for a very traumatic part of your life, very similar, but not exactly, to those with multiple personality disorder act.
> I'm very interested what proponents of the idea, that feelings need consciousness to work, would say about my half-unconscious state.
I’m not one of these proponents, but to play the devil’s advocate: The fact that you can’t remember it doesn’t necessarily imply that you didn’t fully consciously experience it at the time.
How did you jumped to qualia from consciousness? I can see how this is a strong evidence against LLM being conscious, but to my mind it doesn't imply in any way or form that they do not experience qualia.
Or... well, ok, maybe they can't experience if they are not conscious? I see how this can be argued, but I still do not agree. I'm sure qualia is created not by consciousness (I would notice if it was), and I'm sure it is created not for consciousness specifically, it must have some other uses too.
How can qualia be experienced if not from a conscious observer? It's the same question as asking if a tree falls in the woods with no one to hear or know about it.
People have different notions of what “qualia” means. To me, the experience of qualia is the perception of qualities, or of the texture, of inner processings of the brain. Not consciously experiencing them doesn’t mean that these qualities aren’t there, just like not consciously experiencing sounds wouldn’t mean that the sounds aren’t there, and may be unconsciously processed.
I don’t think that “experience as such” makes any sense. Experience is always of something. And that in turn implies that the something that is being experienced also exists independently of it being experienced.
No disagreements with what you said in your first paragraph.
> I'm sure qualia is created not by consciousness
Whether or not qualia is created by consciousness, I don't see how we necessarily can tell one way or the other. We don't exactly have great introspective tools to do such self analysis, not to mention what we think we feel is often illusory/not reflective of reality.
Yeah, I think you are right. "The content of consciousness" and "consciousness" are different things, and I mixed them up, when writing that sentence. I can watch the content, but not other parts that may be producing it.
Moreover, thinking about it, I come to a conclusion, that if I cannot reflect on qualia creation, then it is a (weak) evidence for qualia created by consciousness. I suppose the consciousness is harder to reflect on than other things, hard to map it into states of the content of consciousness. Like, I can reflect on my vision and see some hints on how I get these wonderful pictures, despite it being definitely not consciousness, I can reflect on how I produce or decode language. And to my mind it is because consciousness was devised to reflect on these things, so I could report on my observations to others. But to reflect on itself is a wholly different matter.
I would not deny it. The reason people accept current evidence is, after all, because they can relate to the experience of qualia, even if there's no complete objective understanding of it yet.
I wonder, if they spend their time... lets call it "inefficiently"? Spending so much effort on speeding up interning symbols seems to me at very least inefficient. Whatever rustc spends so much time on, can be anything but parsing. My guess would be that calculating types with all the involved traits is one major culprit, and the other is optimization, because zero-cost abstractions maybe O(1) at the run time, but they are neither zero-cost nor O(1) at the compile time, and there are a lot of them.
So, I wish the author all the best (I believe they have a lot of fun), but I think if their goal to build a fast compiler they are wasting their time.
I'd suggest getting any parser generator, building a parser, and making a working compiler first. Then profile it, and only then start inventing plans for improving performance. Or maybe take rustc and to profile it?
I mean, why to invent super-duper-concurrent-hashtable, if it is not known that it slows down the compiler significantly? Why invent anything? Why not just overcommit memory upfront, making the collision rate vanishingly rare, optimizing the happy path, while letting the collision handling to be as slow as it gets? Will it take too much memory? Really? How much is this "too much"? I don't see any discussions on this, even though this is the most obvious path. With obvious tradeoffs, ofc, but without a measure the tradeoffs can be judged only qualitatively, so you can't decide whether they're worth it.
There are differences between the insider trading and your helicopter example. The theory is the better traders know the reality when making decisions, the better. When oil traders hide information about their reserves, they are working on creating a rift between the reality and the public knowledge. Helicopter is overcoming it. When Trump makes empty announcements that change prices in a purely speculative manner, but before doing this he buys futures, he is just creating instability on the market and he exploiting it. Instability is bad, the whole idea of futures is to deal with the risks stemming from the instability.
Instability is bad, but when the cause of it is market getting new information, it becomes ok: it is bad now, but it is good in the long term. But when the instability becomes a source of profits, when there are incentives and means to create the instability, then long term benefits go away.
I wouldn't give the Flynn effect a lot of weight. The numbers are from IQ tests. No one knows what they measure, they are tuned for a population, for the most of time the Flynn effect had place IQ test scores were used for hiring, school placement, and policy decisions (so Goodheart's Law was at play, how'd you think?).
It is a curious effect, I agree, I'd like to know why it was so, but probably I will not know for sure (I'm a big fan of a scientific method, but I don't believe it is up to a task), and so I personally prefer just ignore it.
I believe they did it all the time. Maybe it was not automated? But they boasted in news multiple times how many coreutils tests they are passing. I suspect that those tests are useless for security, they are more about compatibility or something like that.
> Some, maybe, but if you've decided to rewrite coreutils from scratch, understanding the POSIX APIs is literally your entire job.
Yes, it is. But still such traps in API just unacceptable. If you design API that requires obscure knowledge to do it right, and if you do it wrong you'll get privilege escalation, it is just... just... I have no words for it. It is beyond stupidity. You are just making sure that your system will get these privilege escalations, and not just once, but multiple times.
Psychologists are sure it is. You should get all your traumatic experience and deal with it. You'd better learn how to remember these things without panic attacks or whatever. And the methods they use is replaying the traumatic memories multiple times while controlling the emotional state. The controlled emotional state sticks to the memories and replaces the one that was remembered before.
Well, that's the theory at least. I tried it and it kinda work but not perfectly, it may require some recurrent sessions over time if the effect fades. Though if you practice it a lot, it becomes a habit, an automatic response to traumatic memories, so any memory replay reduces the strength of the memory.
It is like a positive thinking (you get your negative reaction to the memories and replace it with a positive one... well, maybe just less negative), but it is definitely not burying unprocessed memories deep inside your mind.
> the children who drew pictures of the injuries and catastrophe, later showed fewer symptoms of stress and anxiety
I believe it is easier for kids, they are more focused on "here and now", and just replaying a memory in a safe environment has much stronger therapeutic effect than for adults. It is easier for adults to ignore the present safety and to dive deep into their past memories with all the associated emotions, so replaying memories can easily make them worse by intensifying remembered emotions.
Adults have crystallized worldviews, which were probably shaped by their traumatic memories, and it shapes their automatic emotional response, and makes matters worse, harder to change. Children are more fluid, they have more plasticity.
reply