dev_dan_2

Hehe, this was kind of my thesis back 2015; non public unfortunately, but I think the following is save to share: So basically they had the need to persist huge collections of objects that would not fit into a single servers RAM. As for persistence, a key value NoSQL db was chosen (which is a natural fit for a world where everything has a unique identity - I also have some dislike for ORMs because of that, because I saw how beautiful it could synergize, and how ugly some mechanism are that hibernate, EC Framework etc. have to implement in order to brigde the gap between objects and relational dbs).

The core pattern was that the Proxy, which was under Nil and above all other Objects regarding inheritance I think.

It re-implemented `doesNotUnterstand:` by: 1) First loading the actual object from the persistent storage and 2) sending the not understood message to the actual loaded object (which might be able to answer the message instead of calling `doesNotUnderstand:` for every message, like Proxy did.

There where if course optimisations and so on, but that was the gist of it.

What I really like about this system was that it was completely transparent to the sender of the message whether they were talking to a proxy, or the already-loaded object, all while being robust, easy to maintain and so ob. Dealing with collections was tricky though (how much to load at once? what about searching for a particular object?...), and would have been aswell for deeply nested object (which they successfully avoided though because as a SaaS-company, they could model the data exactly to their needs).

I never made that connection, it makes sense though!

Really exciting work and great write up, thanks a lot and all the best to your startup!

`core` instead of `std` is great too!

This will become useful in one of my sideproject where I use bitmaps to speed up pathfinding, exited to try it out!

Oh wow :D Not commenting on the larger picture of why or why not Mistral might be doing that, but commenting on the idea itself: I plan to build my own harness soon, and one of its core tenants was, from the beginning, that it would parse every tool call the LLM attempts and executes them itself. I most cases, that would mean to delegate to one of the tools of the rust coreutils, or making a call to a LSP server.

My reasoning: - That way I will have maximum control over how the LLM can interact with the world (i.e. I can even narrow down which arguments I support/not support) - The monitoring story should be great as well, since every interaction with the outside world happens at one place only - I can add functionality as I go, and can understand how the harness changes and why it does what it does from the ground up

The biggest challenge I foresee is with the LLMs making tool calls and having them react to the results. I also wonder how far I will even get with this :D

Back to topic; "code implemented tool calls" is basically my "oh yeah my harness can do tool calls too" (and everything I need is implemented in one codebase, in... code)

> To the contrary I find why a particular post, especially on an axis like sociology separate from its content, is on the front page to be interesting in many cases. If you don’t that’s fine.

That makes sense, yes, I can understand that!

Regarding the method to do great work, does it not interest you as much as what the sociology angle in general, or did you find this post in particular to be generic/not worth commenting on?

I do not mean this to be lessen the craft of typography (which I am certain that it is!): For me, typography/styling beyond the very basics of bold, underline, italics and so on feels a bit like an aquired taste. Whenever I engage with related contend, I hear claims being made like "A helps to convey X", or "B and C should not mix, else Y!" - and I don't doubt that these claims are true, but I wonder if they are true for a selected few who pay attention to such things, or to statistically average person. My gut feeling would be that there are handful of claims that provably help get a message/intention across to a big group of people due to a mix of convention and how our brain processes what it sees - and at least 3/4 of claims that would not show such effect.

One proxy to try to answer this question would be to look at what successfull advertisements use (not necessarily the ones that win prices, but the ones that drive up sales). Their advertisment media might have other qualities that might offset typography/styling sins, but they are more likely to be intentional about using typography/styling to get their point across.

Commenting on the person that wrote this is not really interesting though - it is in the same category as "I appreciate the fact that 'Paul' has 4 letters and a 2/2 split of vowels and consonants", as far as I am concerned.

Maybe you wanted to make a point about some aspects of the authors biography invalidated some claims they make? That would be interesting, then it would be good to be more specific regarding what you are refering to more concretely, though.

Sure thing! Who would think otherwise?

Being part of the incredibly, exceptionally ambitious group that this essay seems to be targeted at can come in all shapes and forms, I would assume. I am even convinced about it. Music, arts, crafts, social activities, literature, organizing, activism, physics, whatever - it would be very strange to assume that one cannot be highly ambitious in any of those or other areas, I think...

The only thing that matters for being included in that (small, by definition) group is being very ambitious, no statement was being made about which area that ambition should target (at least not by my me). In case you seem to think that PG thinks so: That might be you reading subtext where there is none, PG might have simply refered to the typical Stanford graduate that pitches him a startup idea, nothing more, nothing less.

Yes! The "-ness" implies some degree of spectrum already, and from what I know and understand, one can talk about different degrees/classes of randomness. "decreasing entropy" would work too, I guess (Intuitively: If one would open a bet on a system outcome, how high would the quotas be? If there is some degree of randomness in the system, it would make sense to put money on multiple different outcomes to minimize ones losses. Only if there is no randomness at all, it would be rational to bet all money on the one option we expect).

Another further distinction would be between "true randomness" and "We do not know all input factors yet, hence, the outcome appears random to us". The former is something that genuinely exists in the universe as far as I understand physics (i.e. even if all input parameters are known, it is not possible to predict the outcome reliably, which seems to be the case for quantum physics, which I have the surfacest or surface levels of understandings about), the later boils down to reducing "unknown unkowns".

In practise, this means stuff like using seeds for random number generators, fixing things that look random like day of the week of a DateTime.now(), and so on. Depending on what you do, it might be never possible to get rid of all randomness, but I am 100% convinced that you can always contain it into a well-defined cage, and control everything else down to the last bit.

That also made me think of the opposite: "Increasing predictability" would be acceptable for me too, and a deterministic process/component/system/whatever is one that is at maximum possible predictability for me then.

Either A) he used a rhetorical device, or B) he tried to be relatable and failed (for you, assuming you represent the typical intended audience) or C) he tried to be relatable and succeeded, but you are not part of the intended audience he aimed the piece at. Which is 100%, absolutely, perfectly fine; there is nothing wrong in being in that "very ambitious" group of people Paul Graham is likely refering too. I am not even sure if doing great work is great life advice for the target group, depends on how measures "success at life" or however one would call it.

In any case; I think you will get the most out of the article if you do not pay the relatility too much thought: Even if C) would be true and you (or me or others) are indeed not part of the intended audience, I am sure there are still interesting thoughts in the essay on can think about and potentially apply to ones own life.

Hi HN,

This is my first side project I make public, so I am a bit nervous! :D

If you have any feedback regarding:

- privacy first of all (since I feel most unsecure in this area)

- UX/UI

- concepts / pedagogy ideas

- errors/bugs

- anything else...

... please let me know!

Thank you for your time, and happy hacking!(/infering, if you are so inclined) :)

Dan

> improve determinism

Oh well. I might be to picky here, but how I see things, determinism cannot be improved or worsened, but achieved or not achieved. Or Partially archieved, when analyzing a system that has both components that are deterministic or non deterministic.

There are times to think in absolutes, and when talking about deterministic behavior of technical systems, this is one of them. Join the sith side, we have cookies, and when we say we have cookies, we do.

Seems like a great opportunity to pitch my (optimistic) vision for the relationship between humans and digital media:

- Assume that we cannot trust anything we see on a screen (or listen on a speaker), *based on how real it looks to us*. I am convinced that perception of realness will not be a usefull heuristic for authenthicity anymore. For voice, we are already there as I see it, with images and videos, this is not far off anymore.

- That means that we should be fundamentally sceptic of digital media, until proven otherwise by other mechanisms

- One such mechanism is: Good, old trust between humans. I know you, you know me, we both are really confident that we are truthfull to each other. (You can also through cryptography into the mix to benefit from that trust more independently from time and space, for example by signing an email or a commit).

=> (Hopefull conclusion:) This might lead to us leaving our lives in a more human way again: Trust needs relationships, and relationships need time spent together and nurture.

=> Two nice sideeffects: A) Many methods of blackmail might cease to exist due to plausibly deniability, since many forms of evidence are digital and could therefore be fake B) Propaganda by means of digital media might become less effective, too

Oh wow, instant bookmark for my research folder! I build a hardware simulator, and when I focus on setups with multiple GPUs, this will be a valuable resource.

Love the information density, prose and visual style, thanks for the article!

Looks neat, bookmarked and want to check it out!

Great tutorial, really enjoyed it! Personally, I think languages that can check very much at compile time in combinations with LLMs have a bright future ahead. Additionally, if one wanted to give Haskell a try, Lean4 might be a good language to check out before, as it is more modern and ticks many of the same boxes (Still has some unique features, and the communities quite a lot).

Small feedback: - Great flow, explaination, motivation and so on! :) - Typo: "conext" at the bottom - If you want to keyword-hack a bit, you could introduce a paragraph or too about the role the relationship of Lean4 with LLMs/AI ;)