Hacker News Reader

Absolutely agreed. Brains are bizarre. My mother-in-law once had what we (and her primary physician) believed was a serious flu. The moment we realized that it was much worse than that was when she started calmly speaking total nonsense. She appeared to think it was normal English, but each word was unintelligible, though it used plausible parts of speech. At that moment I picked her up and drove her to the E.R., after which she spent several days fighting meningitis. She came out of it okay, thankfully, but I still vividly remember how scary it was to watch her "talk."

Reminds me of the heavy bertations news caster. She knew what she was trying to say but it was coming out as jibberish and the scary thing is that her brain still recognized the fact that she wasn’t making sense. One of those things that scares me.

Which reminds me, I should mention that for anybody with an elderly relative who experiences a sudden decline in the direction of delirium or dementia: make sure they don't have a urinary tract infection.

I have no idea how that plumbing is connected, but I have seen it happen myself, and apparently it's a not-uncommon experience for people with Alzheimer's: https://www.alzheimers.net/2014-04-03/connection-between-uti...

And yes, it is entirely freaky when somebody you have known for years starts talking utter gibberish and clearly expects to be understood. It feels deeply unreal.

Or indeed other infections. As it was explained to me, elderly people sometimes can't get a fever so the first sign something is wrong can be confusion and other odd behaviour.

In the case of my late mother, it manifested as a fixation on and paranoia about money.

Definitely one to bear in mind.

Talking gibberish is also one of the first symptoms of stroke, but that's well-known.

Doesn't even have to be that serious. My young son went off on a sleep walk once, came in speaking complete gibberish (like a random sort on a dictionary) but in a totally normal rhythm and cadence. After I regained composure and figured out what was going on, and redirected him he want back to bed, artfully working his way around the usual Lego detritus everywhere his path. So weird you can be basically unconscious but still so functional!

Someone I know had a father-in-law with brain cancer that destroyed his ability to speak -- but he could still sing. So, that's how he communicated up until the end.

Makes me think of that buffy episode where they all suddenly sing

So, is tonal language a basic layer below ..?

Singing recruits different parts of the brain.

This is similar to Ozzy Osbourne where he stutters when speaking normally, but has no issues when singing. I've seen reports of this situation with many people.

If you search around the 'net you'll find a number of cases in which a person's principle language has changed after a traumatic brain injury. Some people have even become fluent in a language they only recalled a pidgin vocabulary from before.

Perhaps more common, but still quite rare at about one person per year worldwide, is something called foreign accent syndrome. In which a head injury causes someone with one regional accent to speak the same language with a different accent afterward, often resembling the accent of another region. Examples have included from US to British, US to French, an Oregon woman developed an Australian one, etc.

Sensory changes, changes in personality, and other things can happen with stroke or accidental brain trauma, too.

One particularly interesting case is hinted at in this article, but if you're interested your favorite search engine should turn up many more cases.: https://www.thecut.com/2016/10/the-boy-who-came-out-of-a-com...

> Some people have even become fluent in a language they only recalled a pidgin vocabulary from before.

I wonder if could use this to learn languages more effectively. This kind of changes after a trauma shows that we remember more than we think.

We actively inhibit mixing the different parts of language we know. You can just as well remember actively without destroying another part. What's damaged in most examples here is production, not recollection, anyhow. It helps to talk to people who do not understand your native tongue to actively inhibit production in it.

Correct me if I’m wrong, but I don’t think there are any cases of people with brain injuries speaking languages they couldn’t speak before. And any foreign accent is a case of someone’s speech becoming distorted and coincidentally sounding similar to another accent.

If you are interested in the brain, there are good royal institute lectures on youtube.

The Neuroscience of Consciousness – with Anil Seth

https://www.youtube.com/watch?v=xRel1JKOEbI

The Neuroscience of Memory - Eleanor Maguire

https://www.youtube.com/watch?v=gdzmNwTLakg

Also, any book by oliver sacks is great if you'd rather read books rather than watch lectures.

My mental model is that the brain is a black box and unlike so many other human components, we can't break it open and learn how it works on a mechanical level. We're left with feeding it inputs and seeing what the outputs are as a way to understand how it works.

This means that at best we can have a cause-effect understanding of the brain. Until we can describe the mechanics of the brain with a strong enough model where the resulting behaviours can be predicted, we can't truly understand what the brain is doing.

That's a false dilemma. It is the same with any process. On a functional level, input output correlation is defining. On a mechanical level, we may recognize a structure equal to another black box. Indeed, dividing a structure into separate components is the key technique of learning. The problem is just how highly complex the brain is and that the "output" here is rather different from that of a blood cell or a dna strand. In fact, being based on and perhaps even interacting with those and other building blocks, is what makes the brain so abstract. So speech modeled as the output of dna is necessairly higher up the chain of abstraction than a microscopic ribose enzyme or what. But it appears to as us macro precisely because we don't understand all of it. You have effect and cause reversed.

If we knew where to look better, it wouldn't be a problem. Chances are experts know way better, but there are really few who know enough to see a bigger picture and can guide the search. I'm trying to say they are breaking it open, inserting probes and imaging it working. But as far as the "output" is concerned, everyone is naturally aspiring to be an expert. A lot of unguided superficial theorizing may be informative as well, to great length as it concerns psychology and related fields.

Also brains can remap themselves over time so if pathways are broken, sometimes new pathways can be built.

It's effectively biological version of fuckit.py [1]

[1] - https://github.com/ajalt/fuckitpy

Sounds like shellscript.

IANA neuroscientist, but it also seems that the brain has favored pathways and will follow alternates and recall alternate information sometimes if a primary skill or memory is unavailable. This woman was already fluent in both languages and fell back to German. Other cases have had people show greater skill with a foreign language than they had realized they possessed before the injury, and as they healed their primary language came back and their conscious skills with the other faded.

The most important: Theres no documentation saying how it works, its not modularized, whatever bastard got the idea to create such a system... nice, clean, elegant and consistent abstractions and APIs, please!

The idea of a brain API is both amazing and terrifying.

One could argue that such an API exists. It even understands natural language!

... and then does the opposite of what it’s told.

It's an API. Nobody said the purpose was for it to do as it's told.

It accepts input and responds according to its own training. Disobedience is a feature.

You can argue everything is a feature since the architect isn‘t around to contradict you. ;)

Self-documenting code base, right?

Wildly different localized standards though, and nobody can write a transform that actually works.

Wrappers upon wrappers and glue code ...

Rodney Brooks, robot scientist and iRobot founder jokes that God gave creatures consciousness so he could ask them when the hell is going on with them.

> b) they aren't designed

It does seem like a reasonable statement, but when you take a bit of time to think about it, billions of people disagree with it. A sobering thought.

> so there's no expectation that they should make sense under the hood

What doesn't make sense is that at first you deny a religious explanation, then without skipping a beat you ascribe (what can ironically be described as divine) mysteriousness and inscrutability to a body organ.

This not only doesn't pass the Occam's razor test, but also assumes a defeatist position that "we might never know for sure".

EDIT: I've taken uncharitable tone with this comment, and it may come across as too personal, which wasn't my intention.

actually, if you believe in natural evolution, brains are designed.

just because there isn't any intelligence involved, doesn't mean that the brain isn't a meticulously crafted piece of perfection.

Our brains evolved by having the features needed to keep us alive long enough to reproduce.

Ultimately we, as a species, designed our own brain.

Hydrogen, given long enough, starts to wonder where it came from, and where it is going.

It's probably not, though. Evolution doesn't select for perfection, and many aspects of our evolutionary derived bodies are frankly bizarre.

Take the eye as an example. There's no need for us to have a blind spot. Octopus have eyes almost identical in design, but without a blind spot - convergent evolution meant that both species ended up developing very similar eyes but with one crucial difference: in humans, the retina is wired the wrong way round.

The retina is like a huge array of light detecting cells. In octopus, these cells are connected to the optic nerve on the back side. In humans, they connect on the front side. The blind spot in humans is because of the need for a hole in the retina for the optic nerve to pass through and split and connect to the retina proper.

If there isn't intelligence involved, then it definitely isn't designed. Or crafted, for that matter.

I suppose it could be perfect, by some particular criteria. But it seems unlikely. And in the case of the human brain, it's definitely untrue. The selection criteria for the human brain have been shifting wildly over the last few million years, so it's guaranteed not to be perfect.

Note, for example the long list of cognitive biases that we've so far discovered: https://en.wikipedia.org/wiki/List_of_cognitive_biases

> What doesn't make sense is that at first you deny a religious explanation, then without skipping a beat you ascribe (what can ironically be described as divine) mysteriousness and inscrutability to a body organ.

Are you saying that if we don't understand something, it follows that it must be of divine origin? That hardly makes sense. I don't see the contradiction in the op you seem to see here.

Oh, we're voting on the truth? That's how that works? Please.

You're also wrong about the religious part. Plenty of people who believe in a god or gods believe that humans were created along with the rest of the universe, but many don't believe that they were designed in any useful sense of the word. You can talk to pretty much any scientist who's religious about that. E.g., a Christian astronomer may believe that his god started it all going with the big bang, but does not specifically intervene in the unfolding of creation except at specific points where miracles happen.

> you ascribe (what can ironically be described as divine) mysteriousness and inscrutability to a body organ

I don't in the slightest. I'm pointing at the contrast between my expectations about a) a system that has been expressly designed by minds similar mind to serve me and be comprehensible to them for ongoing work, and b) something that evolved over time with no such constraints. One can eventually come to understand something like that, but I think it would be foolish for an electronics expert to crack open a skull and expect that it would make sense in the same way that they would experience if they opened an iPhone.

> I've taken uncharitable tone with this comment

Yes. And given that I'm talking about my close relatives with brain tumors, I think you've picked a terrible place to do it.

> humans were created along with the rest of the universe, but many don't believe that they were designed in any useful sense of the word.

Genesis 1-27: "So God created man in his own image, in the image of God created he him; male and female created he them."

If you create an MMO and then host it somewhere where it can run in perpetuity and then you move on and never touch the servers again, can you in good faith make an argument that the game wasn't designed by programmers, artists, composers, etc?

So the players might have stumbled into a game (they were born) and believe that the game itself was created, but not designed?

The people you talk about might believe in "a" god, but certainly not the christian god. It doesn't make sense to say that you are a christian, but you believe in it à la carte.

The say they are Christians. Some even teach science at Christian colleges. I'm not sure why you think you know science (and theology) better than them given that it's their life's work.

Also, your argument from that chunk of Genesis is weak. Created "in his own image" could mean quite a lot of things. Imagine that tomorrow humans get the ability to create new universes where they can speed up the flow of time. Pretty much the first thing we'd do is to run a bunch of universes along the lines of our own to see if intelligent life like ourselves turns up.

If we succeed, did we create life? Yes. In our own image? Yes, definitely, as long as your reading of "image" is not extremely literal. Did we design it? No way. And a similar argument would apply if we jumped in further and directed evolution toward particular ends by artificial selection. We would create, but not design.

If, on the other hand, you're very literal about "image", then it requires all sorts of questions. Does your god have a nose? Does it have boogers? If not, why do I have a nose and or boogers? Is he not good at designing in his image? Et cetera.

And regardless, an approach to biblical hermeneutics that boils everything down to one sentence is generally quite weak. It's a big book. Hinging a conclusion about a complicated topic on a single verse is like doing interior decorating while looking through a toilet paper tube: an excessively narrow focus means you're going to miss things, and you're very unlikely to get the whole thing in balance.

minds are much more complicated

It's actually not clear if brains are more complicated to understand than modern smartphones. Modern smartphones are designed using numerous abstraction levels - it makes it easier to design, but harder to understand if you don't know those abstractions. Brain, on the other hand, might have fewer fundamental algorithms/methods. In fact, most of the brain complexity could just be evolutionary quirks, some to make to make it's slightly more efficient, others to make some biological functions easier, etc. In the end, it might turn out that the actual human intelligence is not that hard to understand once we got through understanding all those design quirks.

This might sound ridiculous - after all a human brain looks so much more complex than an iPhone! But if you didn't know anything about transistors, analog circuit design, digital logic design, computer architecture, OS architecture, and software engineering (not to mention the complexity of cellular protocols), you would have a really hard time reverse engineering a smartphone.

I think you hit the nail on the head here.

The brain is perfectly understandable, it's just that we need to learn to understand more things about it before we get there.

We designed everything in a smartphone and can say with confidence that it works a certain way, and even explain why. The brain was designed by evolution, and while we mostly understand what it does, we aren't quite sure of the how and why yet.

> The brain is perfectly understandable, it's just that we need to learn to understand more things about it before we get there.

Perhaps. While the mobile phone might be more complicated than the human brain (I suspect not), that doesn't mean that the human brain is perfectly understandable.

The smartphone was designed by a large number of groups that all built and shared abstractions with the express purpose of these being understandable by other groups. The smart phone's design is purposely built to be comprehensible to the human mind. There is no clear reason to expect that there are clean, human friendly abstractions we can make about all the different layers on which the brain inter-operates.

To analogize to software, it may be like the difference between diving into well documented code that has: clean separation of concerns, standard libraries and design patterns, hardware agnostic implementation, a modern programming language vs code that was coded by a self-taught developer using design patterns that they made up, employs no separation of concerns and relies on hardware timing and interference quirks.

The former may have far more complicated functionality, but still be easier to understand and reason about than the latter because the former was designed to be easy to understand.

No single person understands how a modern smartphone is built. From the camera optics, transistors, OS, wireless systems, AI, etc it simply covers enough domains to be incomprehensible for a single person. What abstraction provides is a means for organizations to understand it.

By comparison DNA is actually a fairly reasonable, program 3 Billion base pairs, but fairly compressible and that’s covering far more than just the brain. But the brain lacks those easily separated chunks, making it harder to decode but not nessisarily more complex.

still be easier to understand and reason about than the latter because the former was designed to be easy to understand.

All those things you mentioned: clean separation of concerns, standard libraries, etc, are only easier to understand if you already know about them. If you have no clue what a standard library is, what it does, or how it fits in the design, then you're faced with more complexity to understand than a code which implements everything from scratch.

You will note that I said minds, not brains. Even once we understand the hardware of the brain, I think we'll have an incredibly long way to go to understand the mind that runs on top of it.

I'll add that the neurologists I've talked with understand a great deal about the hardware of the brain; the organ has been receiving intense study for a long time. But even for them their most common reaction is ¯\_(ツ)_/¯ when confronted with basic questions like "why is he doing that?" and "will he get better or worse?"

You will note that I mentioned both hardware and software components of the smartphone. If you have no idea where one starts and another ends, then all you see is electrons running around in circles, both in a brain, and in a smartphone. Most mobile CPU designers would have a hard time debugging iOS apps, and most iOS devs would have a hard time debugging Apple A12 design. This, however, does not mean that fundamental principles which make a smartphone "smart" are hard to understand: concept of a stored program, high level programming language commands, user interface, etc. But these fundamental principles are buried behind massive design complexities which might not be directly related to the "intelligence" aspect of the phone (computer). I suspect (hope) this also applies to brains.