nemo1618

Huh? I'm not aware of anyone else who defines "deterministic" that way. "Deterministic" comes from "determinism," as in "the effects are fully determined by the causes" -- not "determine" as in "deduce."

LLMs are not inherently non-deterministic. This is a common misconception. You used to be able to set temp=0 and a fixed seed and get the same output every time. This broke when labs started implementing batching, and no one bothered fixing it because the benefits of batching vastly outweighed the demand for deterministic output.

I am hopeful deterministic output will return, though; DeepSeek v4 claims to have implemented "bitwise batch-invariant and deterministic kernels," though I haven't tested it myself.

Sufficiently-developed concentration gives you access to the jhanas, which are extremely blissful states of consciousness. Having reliable access to high valence reduces your need to seek pleasure in less wholesome things (drugs, food, twitter, etc.)

Sufficiently-developed attention gives you insight into how your brain is constructing what you perceive as reality, leading to a reduction in ego, permanent reduction in baseline suffering, and a pervading sense of unity with the rest of the universe.

I'm old enough to remember when companies worth $1 billion were called "unicorns." Now we have a company raising 122 times that? Valued at nearly 1000 times that...?

At least they're throwing consumers a bone via the ARK deal. It's crazy how little AI exposure is available to anyone who isn't already wealthy and/or connected.

yep, I do a simple version of this in Google Sheets. Very useful to be able to "Ctrl-F" your life, especially when combined with Google Maps location history.

If this was a joke, it certainly flew over most people's heads...

This will happen with GUIs as well, once computer-use agents start getting good. Why bother providing an API, when people can just direct their agent to click around inside the app? Trillions of matmuls to accomplish the same result as one HTTP request. It will be glorious. (I am only half joking...)

> But like humans — and unlike computer programs — they do not produce the exact same results every time they are used. This is fundamental to the way that LLMs operate: based on the "weights" derived from their training data, they calculate the likelihood of possible next words to output, then randomly select one (in proportion to its likelihood).

This is emphatically not fundamental to LLMs! Yes, the next token is selected randomly; but "randomly" could mean "chosen using an RNG with a fixed seed." Indeed, many APIs used to support a "temperature" parameter that, when set to 0, would result in fully deterministic output. These parameters were slowly removed or made non-functional, though, and the reason has never been entirely clear to me. My current guess is that it is some combination of A) 99% of users don't care, B) perfect determinism would require not just a seeded RNG, but also fixing a bunch of data races that are currently benign, and C) deterministic output might be exploitable in undesirable ways, or lead to bad PR somehow.

IMO this is one of the best use cases for AI today. Each function is like a separate mini problem with an explicit, easy-to-verify solution, and the goal is (essentially) to output text that resembles what humans write -- specifically, C code, which the models have obviously seen a lot of. And no one is harmed by this use of AI; no one's job is being taken. It's just automating an enormous amount of grunt work that was previously impossible to automate.

I'm part of the effort to decompile Super Smash Bros. Melee, and a fellow contributor recently wrote about how we're doing agent-based decompilation: https://stephenjayakar.com/posts/magic-decomp/

"The steamroller is still many inches away. I'll make a plan once it actually starts crushing my toes."

You are in danger. Unless you estimate the odds of a breakthrough at <5%, or you already have enough money to retire, or you expect that AI will usher in enough prosperity that your job will be irrelevant, it is straight-up irresponsible to forgo making a contingency plan.

I think it's a combination of a) reflexive dislike of any hyped-up tech, mainly due to the crypto era, and b) subconscious ego protection ("this can't be legit, otherwise everything I've built my identity around will be thrown into question").

The best models already produce better code than a significant fraction of human programmers, while also being orders of magnitude faster and cheaper. And the trendlines are stark. Sure, maybe AI can't replace you today. Maybe it will hit that "wall" people are always forecasting, just before it gets good enough to threaten your job. But that's a rather uncomfortable proposition to bet a career on.

> humans still need to have their hands firmly on the wheel if they won’t want to risk their businesses well being

What happens when businesses run by AIs outperform businesses run by humans?

Let's be real. The sky is blue because God thought it was a pretty color, simple as. All this stuff about wavelengths and resonant frequencies and human color perception got retconned into the physics engine at some point in the past millennium, that's why all these epicycles are needed.

Centaurs are a transient phenomenon. In chess, the era of centaur supremacy lasted only about a decade before computers alone eclipsed human+computer. The same will be true in every other discipline.

You can surf the wave, but sooner or later, the wave will come crashing down.

Just wait. In a few years we'll have computer-use agents that are good enough that people will stop making APIs. Why bother duplicating that effort, when people can just direct their agent to click around inside the app? Trillions of matmuls to accomplish the same result as one HTTP request.

This strikes me as a very agent-friendly problem. Given a harness that enforces sufficiently-rigorous tests, I'm sure you could spin up an agent loop that methodically churns through these functions one by one, finishing in a few days.

For those curious about division, I wrote a popular uint128 package in Go that uses one of the standard approaches: https://github.com/lukechampine/uint128/blob/3d2701e8e909405...

Here's one attempt: https://x.com/sigilante/status/2013743578950873105

My take: Any gains from an "LLM-oriented language" will be swamped by the massive training set advantage held by existing mainstream languages. In order to compete, you would need to very rapidly build up a massive corpus of code examples in your new language, and the only way to do that is with... LLMs. Maybe it's feasible, but I suspect that it simply won't be worth the effort; existing languages are already good enough for LLMs to recursively self-improve.

Blackpill is that, for this reason, the mainstream languages we have today will be the final (human-designed) languages to be relevant on a global scale.

Eventually AIs will create their own languages. And humans will, of course, continue designing hobbyist languages for fun. But in terms of influence, there will not be another human language that takes the programming world by storm. There simply is not enough time left.

   The skill isn’t being right. It’s entering discussions to align on the problem.
   Clarity isn’t a style preference - it’s operational risk reduction.
   The punchline isn’t “never innovate.” It’s “innovate only where you’re uniquely paid to innovate.”
   This isn’t strictly about self-promotion. It’s about making the value chain legible to everyone.
   The problem isn’t that engineers can’t write code or use AI to do so. It’s that we’re so good at writing it that we forget to ask whether we should.
   This isn’t passive acceptance but it is strategic focus.
   This isn’t just about being generous with knowledge. It’s a selfish learning hack.
   Insist on interpreting trends, not worshiping thresholds. The goal is insight, not surveillance.
   Senior engineers who say “I don’t know” aren’t showing weakness - they’re creating permission.

I'm so tired bros

This is conflating two things: The stuck, and the suck.

As the author says, the time you spend stuck is the time you're actually thinking. The friction is where the work happens.

But being stuck doesn't have to suck. It does suck, most of the time, for most people; but most people have also experienced flow, where you are still thinking hard, but in a way that does not suck.

Current psychotechnology for reducing or removing the suck is very limited. The best you can do is like... meditate a lot. Or take stimulants, maybe. I am optimistic that within the next few decades we will develop much more sophisticated means of un-suckifying these experiences, so that we can dispense with cope like "it's supposed to be unpleasant" once and for all.

You certainly do not need to play music at the speed the performer intended! There are whole genres (and subgenres) based on this. :) Personally, I have found that slowing a familiar piece down by ~5% tricks my brain into perceiving it as novel again, which helps me attend to it more closely and appreciate it more.

Conversely, I use this "block pattern" a lot, and sometimes it causes lifetime issues:

    let foo: &[SomeType] = {
        let mut foo = vec![];
        // ... initialize foo ...
        &foo
    };

This doesn't work: the memory is owned by the Vec, whose lifetime is tied to the block, so the slice is invalid outside of that block. To be fair, it's probably best to just make foo a Vec, and turn it into a slice where needed.

Years ago there was a YouTuber, "Surveillance Camera Man," who went around pointing a camera at people with no pretense. Frequently the subjects were upset by this and became aggressive, even violent. I believe the intended message was that this is a natural and justified reaction to being surveilled, and yet there is little outcry because public surveillance is largely invisible and/or faceless (e.g. just a CCTV camera mounted on a building, rather than a stranger invading your personal space).

The YouTube account is no longer around, but you can still watch it on archive.org: https://web.archive.org/web/20190220131525/https://www.youtu...

yep, https://github.com/lukechampine/slouch. Fair warning, it's some of the messiest code I've ever written (or at least, posted online). Hoping to clean it up a bit once the bytecode stuff is production-ready.

I made my own, with a Haskell+Bash flavor and a REPL that reloads with each keystroke: https://www.youtube.com/watch?v=r99-nzGDapg

This year I've been working on a bytecode compiler for it, which has been a nice challenge. :)

When I want to get on the leaderboard, though, I use Go. I definitely felt a bit handicapped by the extra typing and lack of 'import solution' (compared to Python), but with an ever-growing 'utils' package and Go's fast compile times, you can still be competitive. I am very proud of my 1st place finish on Day 19 2022, and I credit it to Go's execution speed, which made my brute-force-with-heuristics approach just fast enough to be viable.

Ha, I'm in a similar position, having given up on it after a few months of sips. At what point did it click for you?

AI does not need to be conscious for it to harm us.

Liquids are also removed (gradually). For someone already in weakened condition, I would be surprised if the process took longer than two weeks.

Your passwords should never be visible on screen anyway: They go straight from a password manager into a censored input field.