Comment by SubiculumCode | Hacker News Reader

Comment by SubiculumCode | Hacker News Reader

Parent Full thread

It might start that way, but then they begin to ask, how do Space Invaders and Donkey Kong differ? Ah yes, in space invaders, you cannot move up and down, but in Donkey Kong you can. OK lets create a Space Invaders task where you can also move up and down. Does adding that function make the "Donkey Kong" transistor break the modified Space Invaders behavior? Yes?! So that transistor might not be about Donkey Kong, but for moving up and down. and so on...

An example. Early research which lesioned the hippocampus, lesions brought about impairments on memory tasks...but not all memory tasks. Memory of individual items, or feeling of familiarity without recollection seemed to be relatively preserved. Particularly affected however, were memories involving relations between pairs of items...but not always...those item pairs could be remembered by constructing a story about them, or making one item a feature of another item..so it seemed that item relations that were arbitrary were particularly affected by lesions, and showed more "activity" in neuroimaging studies. and so on. The hippocampus seems to fulfilling a role of binding high-lvel percepts into memory traces for which there is not some lawful/generalizable relation. This was a broad over view..but this goes beyond characterizing a brain region a "donkey kong".

> Does adding that function make the "Donkey Kong" transistor break the modified Space Invaders behavior? Yes?! So that transistor might not be about Donkey Kong, but for moving up and down. and so on...

Only there's no transistor uniquely responsible for moving stuff up and down. There are transistors responsible for particular bits of output of particular machine code commands, like ADD or MOV. But they are commonly used by almost all the code, so the most probable difference between code triggering the error and a code that's working correctly - would be "how big values we're working on", and if the values are in the "correct" range - that transistor will make a difference. It very well might be that x coordinates are big and y coordinates are low, so it's working for y but not for x. In that particular level of that particular game, assuming the memory was in a particular state before you started that game.

The problem lies in trying to assign too high-level meaning to stuff that works on much lower level of abstraction. It's very similar to alchemy or astrology. Searching for correlation between unrelated events and basing elaborate theories on that.

If that transistor was not about moving up and down, you'd have to explain why lesioning that transistor seems to break games with up and down motion. As what you say is correct, scientists would soon find other tasks it breaks that do not appear to have up and down motion. You might move to questions about complexity of objects, coding of directions, etc. Theories would emerge and begin to be related to emerging understanding of archtectures, transistor function, controllers. It evolves.

Aside from that, I object to the view point more generally. Can one not understand something about how cultures work without knowing how brains work? Can one not know something about how an ape behaves without understanding their gut-biome? I just disagree with the view that the only worthwhile understanding of phenomena is from the bottom...that there is some true level of description at which something must be understood.

But you could have the "same" program break under different conditions, because you can have two different but equivalent programs.

Consider the following pseudo code:

    i <- 0;
    total <- 0;
    while(i < 10) { 
        total <- total + i
        i <- i + 1 }
    output total

This program could be done using a range of different instructions.

At the most naive it would actually use registers and jump and compare instructions to do the loop.

A different approach would do "loop unrolling" to avoid any jump instructions, instead applying sequentially the addition and comparisons.

The most aggressive approach would simply replace this whole program with

    output 45

Without understanding the architecture and program preparation, what hope is there to map functionality to underlying behaviour at the transistor level?

> Can one not understand something about how cultures work without knowing how brains work? Can one not know something about how an ape behaves without understanding their gut-biome?

Absolutely this. An explanation of evolution in terms of atomic interactions would lack some of the explanatory and predictive power we gain at the higher levels between bio-chemistry and phenotype. I think part of good science (and FWIW software design) is choosing the correct abstraction level, the one with the best predictive, descriptive and explanatory power. Sometimes these will be different abstractions or models that have different affordances.

In the case of the OP I think it's easy to underestimate how hard it is to generate a bottom up explanation of a processor given that we already have such an explanation from its status as engineered artefact. The answer is simple if you know the answer.

Agreed. The iomportant thing to understand about neuroscience is that it is being attacked at multiple levels of decscription (synapse,dendritic trees, whole neurons, small groups of neurons, cortical columns, networks of cortical columns. Then we try to bridge understanding across levels of description and all try to relate it to behavior on one way or another.

Quite right. The movement toward a reconciliation between various levels of abstraction and observation is one (perhaps the greatest one) of the ways we know we scientists are "getting it right".

Peter Watson's book "Convergence" is a delicious description of this broad phenomenon in scientific investigation(s).

> Can one not understand something about how cultures work without knowing how brains work?

Yes, you can, but that's psychology or sociology, not neuroscience. The goal was to understand brain, and be able to replicate it or modify it. Understanding just the behaviour is like understanding that pressing "fire" will shoot in Space Invaders. It's something, but you don't need to dissect a CPU to know that, and it doesn't move you closer to creating your own CPU or modifying this one.

> the only worthwhile understanding of phenomena is from the bottom

That's misrepresentation of my point. I wrote that in the CPU example neuroscience tools failed because of missing the abstraction levels between transistors and behaviours.

> there is some true level of description at which something must be understood

There is optimal level of description, yes. And notation makes a huge difference.

Neuroscience works at a number of levels of description or size. Neuroscience which focuses on synapses Neuroscience that focuses on dendritic tree or whole neurons, Neuroscience that ask questions of small groups of neurons, and Neuroscience that focuses on large scale networks. Most try to relate that function at that level to behavior in some way and to bridge the gap between levels of description.

When we try to understand deep learning, do we try to understand it at the level of individual transistors, at the level of a 'neuron', at layers, or at the level overall design? Well it depends, right?

I'm not a scientist, but it seems the various groupings of neurons yout enumerated are still hardware level, and not the software level.

Maybe there really is no software level between hardware level and behaviours, but it seems unlikely to me, because I can think about abstract ideas using that brain.

Instead of assigning too high-level meaning to stuff that works on a much lower level, it might be the opposite.

The reason why you can say that one transistor doesn't handle moving up and down is because we can engineer and economically produce systems with enough spare resources and performance to implement general computing devices capable of running arbitrary programs.

If instead you were playing say... a very early arcade game, you might very well actually have the 'X coord' register implemented in a fixed discrete component.

One of the assumptions baked into biology is that by the time you're talking about cells, you're talking about implementations of behavior subject to reasonably strict constraints (plus hilarious amounts of path dependent evolution), with the assumption that things are more like old school analog/digital electronics where every component counts, versus modern integrated processors.

That's not to say that they have the perfect analytical techniques for solving problems in their domain space, or any domain space.

Reply on news.ycombinator.com