This is promising. I suppose these funds will also help address the vulnerability of the US power grid to cyberattacks, due to its heavy reliance on a small set of critical nodes.
>When you have a big concept in the back of your mind, it becomes a filter that everything you experience runs through
This is so true, to the point where sometimes you feel possessed by the concept and not the other way around.
The concept harvests your human experience to make its way into the real world.
It defines your notion of signal and noise, it restlessly samples patterns in the real world that might help its incarnation.
Sometimes due to predisposition, part of the concept is hardware implemented which means you never really get to experience what it's like to live without this concept driving your life.
The concept carefully arranges your dreams, strikes you with overwhelming visions that feel more real that your clearest memories and skilfully crafts your personality.
For those who feel as though they are concept integrators, do not allow concepts to mistreat you.
They don't belong to our world, they don't care about the totality of human experience, they operate on timescales far greater than that of our precious lives and our civilisation as a whole is still far too primitive to bear their throughput.
>I would start by simply putting everyone through a course in deductive reasoning at the earliest age possible
Indeed. This would help ensure people's brains' transition function is stable enough to perform faultless computation.
We forget that our brains aren't wired for exact computation. They're wired to perform approximations of computation that are good enough for survival.
As a result, you end up with myriads students who go through the school system via memorization and emergent fuzzy computation.
They reach an adult age without possessing the cognitive tool-set to grasp the subtleties and nuances of the world they live in.
The fact that such people are also preyed on by charlatans, ad companies and politicians(intersection of charlatans and ad companies) obviously doesn't help.
>research and quote all the works done over the last 10 years or so by researchers in this space!!
I totally understand your point and appreciate you linking those resources, however I think it's important to remember that the author's post is from a personal blog, not from a scientific journal or arxiv.
Perhaps OP would've never posted this if he felt that his "contribution" wasn't novel enough.
Additionally, there's a chance that the wording and tone the author used might speak to people who found the articles you mentioned opaque(and vice versa, obviously).
If the author, feeling the urge to write something up, had looked very hard for "prior work" instead of following the flow of their insights gained through experience, perhaps they would've felt compelled to use the same vocabulary as the source, which has its pros(forwarding instead of reinventing knowledge) and cons(propagating opaque terms, self censoring because of a feeling of incompetence in the face of the almighty researchers).
That's one of the great things about blog posts: to be able to write freely without being blamed for incompleteness or prior art omission.
On a different note, I think this may also highlight the fact that the prior work you mentioned isn't easy enough to find. Perhaps knowledge isn't circulating well enough outside of particular circles.
I was thinking that perhaps walking the readers of our code through our architectural decisions(and not just through what our code does) is a good way to lessen their cognitive load.
This helps identify decisions that have been taken to look smart or because the foie-gras the author ate on that day went down really well with the Chardonnay and made them feel extra stylish.
This also helps us understand how well we know the tools we're using versus how much we do simply through pattern repetition.
"23. Has anyone ever compared how long it takes to reach a workable level of understanding of a software system with and without UML diagrams or other graphical notations? More generally, is there any correlation between the amount or quality of different kinds of developer-oriented documentation and time-to-understanding, and if so, which kinds of documentation fare best?"
This is such an important question and it's just the tip of the iceberg of a very deep problem that is rotting our software systems. We are absolutely pathetic at dealing with complexity and we actually enjoy complexity. We don't tackle questions such as 23. ANYWHERE near as seriously as we should.
Developers overestimate their mental bandwidth which leads them to pompously build over-complicated tech stacks despite only having archaic tools to mitigate and navigate their complexity.
Companies don't need to hire more devs to deal with their complex software systems, they need better tools to navigate their software systems. But because companies don't truly value their money and devs don't truly value their time, we end up in the situation we are in now.
We should have hundreds of companies investing on initiatives akin to Moldable Development[1], instead they play the following bingo:
1) let's just hire more devs and hope to land on a 10xer
2) let's build our own framework
Additionally, we overvalue specialization. By overloading developer brains with complex tech stacks, we encourage a culture of specialized profiles who find solace in trivia. Doing so, we limit cross-pollination and stifle true innovation. This attitude is actively killing-off thousands of valuable ideas.
Every second, there's a coder out there who thinks of something wild, which requires very specific tools from different fields and finds out that the people who built such tools couldn't be bothered making them accessible under a sensible time-budget to people outside of their niche/ivory tower. So the dev either drops the idea or gets sucked up into a niche.
This is tragic, but hey look! We have a new (totally not low-hanging fruit that could be predicted 10 years ago) Generative Model, WOW! "What a time to be alive"!
It's clear that the next frontier is to have 3D-space instead of image space transitions. Language itself is very static and action verbs are not enough to specify scene dynamics. I suppose we would need:
A. an enriched version of natural language that refines the dynamic processes that occur in a scene
B. a data set of isolated processes labeled in the language described in A.
I've had a hard time finding ongoing work on A. and B, perhaps it isn't much of a priority for research groups.
>Now try properly aligning with an "entity" that thinks nothing like you, has a vastly different set of experiences, doesn't know how to count, and has no real concept of physical mechanisms (and more).
The biggest issue with alignment is that humans don't really know what they mean nor want in the first place.
Yet, we train these networks to produce high quality images, whose resolution is necessarily higher than the resolution of the fundamentally ambiguous human input.
>it can’t understand the underlying realities in which its paintings are based and generalize those into novel situations, styles, or ideas it has never seen before
As long as we keep putting a greater emphasis on the final output than we do on the process that led to the output, this "limitation" won't be a factor.
Sometimes that point exists in a region of concepts that isn't touched by anyone directly, so you have no clue what its exact location is, hence you can't concentrate your energy on it.
What you can do, however is identify the conceptually closest and well understood points. These points may end up being in vastly different fields, each requiring vastly different ways of thinking and/or computational paradigms.
_IF_ you're resourceful enough to focus your attention on all of these neighboring points "virtually" at once, then you have a chance to interpolate the position the original concept and finally concentrate on it.
>Could we do more to make programming an accessible activity to the 99% of people who don’t program, by putting greater emphasis on intuitive physical analogs?
I think this would have to be done carefully.
Being able to articulate your thoughts and compute without resorting to physical analogies can greatly reduce the processing load on your brain. This is partly due to the fact that analogies always come with a baggage of extra features that are unrelated to the concept you actually care about.
Perhaps a good way to overcome this would be to carefully pick a set of analogies for a concept,then identify what about these multiple analogies helps you understand it. It would also help you abstract away the additional semantic load that comes with the chosen physical analogies.
I am however, a strong believer in the power of metaphors to democratize complex craft.
Music composition per example, can be made more intuitive and accessible by thinking of sounds events in terms of their physical analogues. In the case of sounds however, the extra semantic fat may very well be marbling on wagyu beef.
In fact, it would help us think of musical composition in a more... compositional way, i.e. combination of parts with different semantic properties that when combined following certain rules, generate an expression with a particular meaning.
Can seasoned Blender users can visualize and compose Geometry Node graphs in their minds? It seems to me that it would be pretty hard, simply because of their clunky design.
In my opinion, the only thing missing from Blender's amazing Geo Node system is icons, i.e. something compact and symbolic that one can store in their mental space as they would a word when composing a sentence.
I do also feel like a top down flow is more conductive to "process-oriented" or procedural design. I've done my fair share of procedural visual scripting with nodal systems such as UE4's Blueprints, PopcornFX or Blender's Geometry Nodes and none of them feel as "natural" as Houdini's.
It almost feels as though vertical lines occupy less space in my mind than horizontal lines do. Perhaps it's because our brains are hardwire to deal with lots of top down structured environments (plants, tree-trunks, other Sapiens).
" “Music is really some kind of a magic,” says Tchernichovski. “It has no meaning. And yet, we like it, and we engage in it, and it gives meaning to our lives.” "
I think of that "magic" in music as more of a "protolanguage". It's implicitly spoken, non-symbolic and because we experience very similar realities, it's relatively universal.
Outside from noise the natural sound patterns that surround us are governed by rules (ex: Dynamics of physical interactions[0], linguistic structure[1], and behavioral rules).
If instead of sampling the sounds themselves, you sample the rules that generate the patterns, and you highlight the causality relationships.
Perhaps, if in addition to this you associated linguistic and behavioral dynamics to the internal (emotional) states of agents, you'd be able to explain (and make use of)the universality and suggestiveness of a good chunk of musical patterns.
You should take a look at Denis Villeneuve's Arrival(2016) if you haven't already. It explores this concept in a very peculiar yet thought-provoking way.
That aside, I think that If in the future somehow manage to make effective use of VR/AR for purposes other than entertainment, such virtual environments could be used to simulate suchevents. It would be reminescent of today's Fortnite events, but with much more serious repercussions.
However one should keep in mind that such events could have traumatizing and demoralizing effects on our civilization. Although temporary these effects might constitute a major setback.
The following article from early 2022 gives a thorough yet concise overview of the subject: https://semiengineering.com/power-grids-under-attack/