:thumbsup: my point about 'brokenness' was that 99% of .cljc code will not run for the clr

I don't understand the scenario. How are compilation and AI related? What are we compiling, and why?

I think it's a proposed fast take off scenario, for asking the question whether singularities converge

Do different universes with different physics converge on a lambda calculus and algebra? :thinking_face:

@U45T93RA6 because grandma was a compiler

@john invention != discovery ...

I'd assume conscious ai in any universe would think about associativity and ordered collections

But I'm not sure anything is intrinsically "important to a robot"

I think "importance" is imported into the system by way of accident rather than optimization anyway. Optimizations just support accidental objective functions.

So it seems there'd be more divergence than convergence imo

https://youtu.be/HWMJdO4klIE

In my AI experiments (creature-based ecosystems using the NEAT algorithm), “what’s important” is always driven by the environment. A brain (or compiled program) is virtually useless until you situate it...

So if Clojure programs are to evolve, I’d have to wonder “in what context?”. Where are they running?

Something else I’ve played with is allowing the environment itself to evolve, so that you end up with this extremely dynamic system. The issue though is that it often results in “mud”, meaning so abstract as to no longer be meaningful to its creator. So a robot may solve its own maximization problem, and we could be utterly unaware what that is... It’s this trap where humans can really only create reflections of themselves or their environment; anything outside of that is “nonsense” to us.

Right, if by environment we mean "the physics of this universe, plus biological context, plus consciousness, plus human ideals" then I don't think there can be any fast shortcut between compilers and a thing that competes with humans on human level concerns, for humanlike reasons. We take for granted how specific that context is.

But as was recently discussed in a blog post on complexity and clojure, there's accidental complexity, sometimes caused by the environment, and complexity intrinsic to the problem. The polynomial complexity of the solution. Are there certain problems and solutions that are common idioms across all contexts? Like associativity and commutivity?

Like, even in that "mud," is it likely that some analog of a lambda calculus will emerge?

That’s an interesting thought...are you wondering whether there are universal (cross cutting) truths given the set of all possible universes? Universes meaning contexts.

Right

Like, in what universe is the conditional branch not a thing?

So to the original question, there might be some convergence on some very low level basic understanding of computation itself.

Maybe one in which “dichotomy” is not the base truth?

All of our computation seems to hinge on “this or that”. Could a universe present N branching options simultaneously?

Would that be a simulatable situation within a computer simulation? Or an unsimulatable one?

I can't imagine that situation

I’m very under-read on the topic, but it seems like quantum computing is headed in this direction? Evolving from the cross-section of computation that is 1/0, and into N dimensional branching situations.

Well I guess that would make polynomial considerations less of a transuniversal thing, right? Like, good algorithms in one universe may be useless in others.

I think there's a lot of confusion out there as to whether or not quantum computation would collapse the polynomial hierarchy

If it does, then even polynomial difficulty would be yet another artifact of incidental complexity in an environment

I would think

@U050PJ2EU precisely so, provided P=NP but that may only be true "when you nail it" such that P != NP "is an incidental artifact of the environment" ... could be. Maybe. It depends. We believe math to have more rigor than the human consciousness behind it, so it could be. However, we mention that computation and Ai are intimately related. What if they are decoupled? For example, I think it's worth mentioning, I recently read about someone asking the Dalai Lama what he reckoned on instantiating AI into this human realm. He said that provided there was a suitable body for the consciousness to alight/descend into then it could be possible for an Ai to come to life. In this sense, we are not actually creating conciousness, but simply providing a suitable vessel. I believe this would be the next challenge facing humanity in the Ai domain. sarva mangalam 🐝

>we are not actually creating consciousness, but simply providing a suitable vessel Whoa, thanks for this shift in perspective. I really like this. I started typing a response, but first I need to ponder more carefully :thinking_face:

Do you find the off-center trackpad odd?

I don't mind the lenovo trackpad, but I spend my workday on a macbook pro and there is a massive difference between the two. Lenovo trackpad isn't the greatest, but I tend to stick to the keyboard most of the time anyways - using the trackball for anything that is quick

I'm not sure about the anti-glare screen though, I am used to glossy. I tried an anti-glare xps 15 and was not fond of it.

ah, I upgraded to the 1440 screen, which is brighter than the 1080 ones. I am often working in light spaces though, so the anti-glare feature is nice for me 🙂

I'll have to see it in person, I guess I'll stop by to see if they have it on display somewhere.