I've been reading and reading about DeVault for more than a decade now. If I can point to one person on the internet and definitively say that their today's version is better than a decade ago, it would be him. (Yes, I can say that he appears to have improved better in this time than I myself have, which can be interpreted in a more than one way).
In fact, Andrew Kelley, whom I respect fair bit, also chose to stand behind redict, Drew's fork of redis with similar observation.
People change over time, some of them for the better, and I personally like to give them a chance. Some of Drew's opinions and expressions are still a bit much for me, but that is just us both being human.
> a better build system is sorely needed - one of the things that Go / Rust did right
Honestly there are only two reasons I wouldn't pick up Java for personal projects, difficult to build single executable (Graal is still very un-ergonomic), ridiculous build systems.
I can kind of live with former, but Gradle is so very extraordinarily terrible that I don't know where to begin. Problem is, it solves some real problems (in extremely bad way) that people keep using it.
I long of a cargo-style revolution in Java world. (No, the newly popped up alternatives haven't really cut it so far)
I mean, sure, a good, detailed commit message is perfectly fine to me in place of the prompts / a session distillation. But I am not holding my breath for vibe-coders to properly review their code and make such a commit message. But, if they, do, great! No need for prompt / session details.
> How does the CCP avoid empire-building, institutional rot and general bureaucratic paralysis?
Oh they don't! In exactly the same way US didn't. Right now, a lot of factors have put enough tailwind into Chinese economy and the inertia is a bitch to retreat, as can be seen with US itself. These tailwinds are strong enough that they lift everybody up, even considering the corruption taking its share.
> all come out at “question mark in a box” (Chrome, Edge) or “codepoint hex in box” (FF) on the old Win10 box that I'm currently trying to retire. The come out find on a similarly default Win11 setup.
This is pretty funny to me, because on plain ol' Firefox on NixOS everything looks just fine!
We've come pretty far from the days when things were randomly broken on Linux..
Firefox on Fedora 43 here, no joy, and indeed, randomly broken on Linux (too).
[edit] Okey, so installing the complete Google Noto Fonts family resolved this issue. But I still don't know if relying on a script with almost no font support is any better than what the OP did.
Empires always fall from within. It was inconceivable for a young me to ever think of day when MS Office would be unworkable. Advance couple of decades and MS 365 Copilot is just the thing that just doesn't work. Not because somebody exploited a bug and created unviewable doc, but because MS decided to pile on bugs while leaving old ones in..
> Just don't complain when large corporations copy your work one day with no legal recourse.
To be fair, that is the schadenfreude. Large corporations have been copying works of little people for ages. They only started crying about 'IP theft' when someone bigger (China) started doing the same to them, and to make it worse, most of the corps willingly handed the IP over because they wanted cheap exploitable labor.
Hallucination is all an LLM does. That is their nature, to hallucinate.
We just happen to find some of these hallucinations useful.
Let's not pretend that hallucination is a byproduct. The usefulness is the byproduct. That is what surprised the original researchers on transformer performance, and that is why the 'attention is all you need' paper remains such a phenomenon.
I wish people who take this stance would seriously reconsider their take on how hallucinations are defined and how unhelpful it is to conflate hallucination with generation from a probability distribution. I appreciate OpenAI publishing articles like this because, while the parent comment and I may have to agree to disagree on how hallucinations are defined, I can at least appeal to OpenAI's authority to say that such arguments are not only unhelpful, but also unsound.
You're going to get a lot of pushback on the idea of taking the definition of hallucination seriously. Calling fluently stated bunk "hallucination" feels cynical to begin with. Trying to weave a silk purse out of that sow's ear is difficult.
I don't know what you mean by hallucination here; are you saying that any statistical output is "hallucination"? If so, then we are also constantly hallucinating I guess.
There doesn't seem to be a particularly consistent definition of what "hallucinate" means in the context of LLMs, so let's make one that is in line with the post.
"Hallucination" is when a language model outputs a sequence of tokens comprising a statement (an assertion that is either true or false) that is incorrect. Under this definition, hallucination is clearly not all that an LLM can do.
An easy way to avoid hallucination under this definition is to respond with something that is never a statement when there is a possibility that it can be incorrect; e.g. "I think that... I don't know...". To me, this seems to be what the authors argue. This has always seemed pretty obvious to most people I've spoken to (hell, I've reviewed grant applications from years ago which talk about this), so I'm not sure why it took so long for the "frontier" developers to actually try this.
In fact, Andrew Kelley, whom I respect fair bit, also chose to stand behind redict, Drew's fork of redis with similar observation.
People change over time, some of them for the better, and I personally like to give them a chance. Some of Drew's opinions and expressions are still a bit much for me, but that is just us both being human.
reply