Hacker Newsnew | past | comments | ask | show | jobs | submit | metabrew's commentslogin

I tried the chatbot. jarring to see a large response come back instantly at over 15k tok/sec

I'll take one with a frontier model please, for my local coding and home ai needs..


Absolute insanity to see a coherent text block that takes at least 2 minutes to read generated in a fraction of a second. Crazy stuff...


Accelerating the end of the usable text-based internet one chip at a time.


Not at all if you consider the internet pre-LLM. That is the standard expectation when you load a website.

The slow word-by-word typing was what we started to get used to with LLMs.

If these techniques get widespread, we may grow accustomed to the "old" speed again where content loads ~instantly.

Imagine a content forest like Wikipedia instantly generated like a Minecraft word...


Yes, but the quality of the output leaves to be desired. I just asked about some sports history and got a mix of correct information and totally made up nonsense. Not unexpected for an 8k model, but raises the question of what the use case is for such small models.


8b models are great at converting unstructured data to a structured format. Say, you want to transcribe all your customer calls and get a list of issues they discussed most often. Currently with the larger models it takes me hours.

A chatbot which tells you various fun facts is not the only use case for LLMs. They're language models first and foremost, so they're good at language processing tasks (where they don't "hallucinate" as much).

Their ability to memorize various facts (with some "hallucinations") is an interesting side effect which is now abused to make them into "AI agents" and what not but they're just general-purpose language processing machines at their core.


Would be nice to point this at (pre-LLM) Wikipedia and fill out Wikidata!


You have a misunderstanding of what LLMs are good at.


Poster wants it to play Jeopardy, not process text.


Not sure if you're correct, as the market is betting trillions of dollars on these LLMs, hoping that they'll be close to what the OP had expected to happen in this case.


The market didn't throw trillions of dollars to develop Llama 3 8B.

What GP is expected to happen has happened around late 2024 ~ early 2025 when LLM frontends got web search feature. It's old tech now.


The GP’s point was about LLMs generally, no matter the interface. I agree that this particular model is (relatively speaking) ancient in AI the world, but go back 3 or 4 years and this (pretty complex “reasoning” at almost instant speed) would have seemed taken out of a science-fiction book.


I don't think he does. Larger models are definitely better at not hallucinating. Enough that they are good at answering questions on popular topics.

Smaller models, not so much.


Care to enlighten me?


Don't ask a small LLM about precise minutiae factual information.

Alternatively, ask yourself how plausible it sounds that all the facts in the world could be compressed into 8k parameters while remaining intact and fine-grained. If your answer is that it sounds pretty impossible... well it is.


Did you see the part in my original post where it said "Not unexpected for an 8k model"?


Oh I saw it, you still have a fundamentally flawed comprehension of LLMs.

The size of the model does not factor as tiny models can use Internet to fetch factual information.

But you think they are accurate repositories of knowledge, even though it's physically impossible unless lossless infinite compression algorithms exist (they don't, can't and won't).


I think you're overestimating your ability to assess what others think or comprehend.


Reminds me of that solution to Fermi's paradox, that we don't detect signals from extraterrestrial civilizations because they run on a different clock speed.


Iain M Banks’ The Algebraist does a great job of covering that territory. If an organism had a lifespan of millions of years, they might perceive time and communication differently to say a house fly or us.


:eyeroll:


Yeah, feeding that speed into a reasoning loop or a coding harness is going to revolutionize AI.


this is correct :)


Did this with policy based routing in my opnsense (pfsense) router a couple of weeks ago. egress via a specific tailscale exit node for a list of domains including Imgur.

Also browsing Minecraft mods/shaders was my motivation ha.


I would love to pay for youtube premium, but i have a google workspace/apps/own-domain/whatever the hell they call it now account, and loads of stuff (like youtube premium) isn't supported.


I had that issue. I just created a new account solely for YouTube


Same. It’s wild that Google can’t sort that out


Make another account.


Will definitely need to tone down the American Corporate Alacrity for the UK market..


Just get rid of it all together. I want my device to sound dry and factual like the ship computer in Star Trek, not emotional and... moist... like the lovechild of a Youtuber and a SV startup bro.


Well, you're not the only one who wants things. I wouldn't mind some Her style interactions in some of my assistants, not everything needs to be bone dry.


Many opponents to the bill have been very cagey about their reasons for opposing it, and eventually admit it's for nebulous "religious reasons".

Personally I'd like the right to die with dignity if I were unfortunate enough to find myself facing horrible, imminent, certain death.

I'm glad it passed and I hope it makes it into law.


I dislike these proposals because they always eventually encourage the erosion of help for people who very much want to live. I don't think the end of life is very dignified, but I don't trust the government to ever decide to spend money helping people in pain reduce their pain, instead of shrugging and saying, "Well, you can choose to die" We're seeing this in Canada and it was utterly predictable


How much money should governments spend to keep a patient alive one more month? If you've actually thought this through then you should be able to reply with a specific number. Resources are limited so anything spent on caring for those patients can't be spent on other priorities. And don't try to weasel out of the question by claiming that we need to raise taxes or cut military spending or whatever; no matter what we do with other budget line items the answer always has to come down to a specific number.


In Canada they have proposed assisted dying for young people suffering from depression. What better way to harvest organs and save money than by making unproductive young people want to die?

If this was seriously restricted to people facing severe debilitating injuries or pain I might be more in favor. But you know it won't stop there.


Does the Bible say you shouldn't end your life if you're dying of something incurable and terribly painful?


Regardless of what a bible says it is indeed the view of many religious people.


They probably change their mind pretty quick if they witness a loved one go through endless pain for no reason. Some of them, I'd think. Others, I'm sure, would tell them how much wonderful the gates of heaven will look when they get there.


Yes. Every Jew and Christian, until recently, interpreted thou shall not kill (more accurately translated as murder) as a ban on killing yourself or another with the intention of killing yourself.


Thats odd. Anyway seems like that is a possible reading but not all possible readings.


If every Jew and Christian until like 100 years ago think that was the correct interpretation then that is the canonical interpretation. Calling in a possible reading is just not correct.


It would seem that many exceptions were carved out to allow Christian led crusades, pograms, etc. Jewish military history is extensive and embraced more than simple stern words: https://en.wikipedia.org/wiki/Jewish_military_history.


I am not sure what your point even is. What is typically referred to as killing (thou shalt not kill) is referring to murder / unjustified killing. This is how Jews and Christians have interpreted it for almost the entirety of their existence. It is only in the more recent years that a small minority have disagreed.

There is no exception because it is not talking about just killings only unjust ones.

Wars, and killings within them, can be just (look up just war theory). The point of a war, at least from a Christian perspective, is not to kill people, but to stop an invasion, protect innocents or the like. If it isn't doing that then the war is not just. If you are wanting to kill, rather than protect innocents or whatever then you are fighting for the wrong reason. This is what is known as the principle of double effect.

The Crusades were intended to stop the Muslim invasion, regain Christian lands, prevent the fall of the Byzantine Empire, allow for safe travel of pilgrims, etc. Those are considered just things to protect. (Obviously not all people in the Crusades had the same motives, but that is on them not the war in general).

Also, just because Christians, Jews, or whoever do something, doesn't mean it is inline with their faith. People justify bad things, ignore teachings, don't actually believe, etc.

To give a more modern example and one relevant to the thread, jumping on a live grenade is not considered suicide because you are not attempting to kill yourself, but to stop others from being blown up. You know you are probably going to die, but since you are not intending to kill yourself it is not an unjust killing. Jesus said "Greater love has no one than this, that someone lay down his life for his friends".


The point is a simple one, in the space of a few comments you've moved from:

    Every Jew and Christian, until recently, interpreted thou shall not kill (more accurately translated as murder) as a ban on killing yourself or another ..
to

    (paraphrased) except for when justified.
meaning that there has been considerable latitude in the reading and interpretation of what in modern english is stated as "Thou shalt not kill"

There will be some who interpret:

    Jesus said "Greater love has no one than this, that someone lay down his life for his friends".
as making suicide acceptable so as to not place a burden on friends and family due to { reasons }.

There are many Christians, Jews, and Muslims in this world and many interpretations of the many and varied texts and translations.


I may not have been as clear as I should have been.

I was making a distinction between killing and murder. All murders are killings, not all killings are murders. If you kill somebody it may be acceptable or it may not be. Killing somebody in defense of another is fine so long as you are intending to protect the innocent not to kill. You may end up killing, but it would not be murder and as such would be justifiable.

Bringing up the latitude and interpretation is a major problem some Christians have, but this is a modern problem. The Church, historical, had always believed there needs to be people involved to help guide people. The Bible itself makes the point in Acts 8: "So Philip ran to him and heard him reading Isaiah the prophet and asked, “Do you understand what you are reading?” And he said, “How can I, unless someone guides me?” And he invited Philip to come up and sit with him."

You are highlighting a reason why somebody shouldn't just read the Bible and assume they know what it means. This is a problem that only applies to a portion of the Protestant churches not the remaining Protestant churches or the Churches that existed before the reformation (Catholic, Orthodox, etc).

You bringing up English translations is why I was making it clear that murder was a more accurate way of understanding the decree against killing. I knew if I didn't specify that it was only murder, somebody would ask about war. Apparently I wasn't clear enough.

Can you find a single prominent Christian or Jewish scholar, theologian, saint, etc who lived 1000 years ago who believed that Jesus was allowing suicide?

Like I said, this is a modern problem because some Christian churches have gone off the rails and have rejected the traditional views held by the Church. They are clearly wrong. Using them as justification for something is ridiculous. Imagine if somebody said Stalin / Mao / whoever was a communist. Communists are atheists so that means atheists can kill people like those people did. You would argue that those people do not represent an accurate description of atheists. Just as you would reject that argument against atheists, I reject your argument against Jews and Christians.


> Can you find a single prominent Christian or Jewish scholar, theologian, saint, etc who lived 1000 years ago who believed that Jesus was allowing suicide?

The debate on whether jumping on a grenade to save others is suicide or self secrifice, et al has passed you by?

Is it your claim that not a single prominent Christian or Jewish scholar / theologian has wrestled with that conundrum and if they had they all fell in line in unison?

> because some Christian churches have gone off the rails and have rejected the traditional views held by the Church.

There's cetainly less church sanctioned burning of Catholics and Cathars alike .. I don't believe there's a uniform monolithic corp of "traditional views" prior to modern times .. even the implication of a singular Christian "the Church" appears to be an ahistoric over simplification.


bravo for trying. its bizarre to me that we are at a point in time that someone could even imagine that suicide, esp assisted, would be permissible in Christianity.

the protestant view dominates "public Christianity". even Catholics will happily tell you that "yeah we believe in this, but I certainly dont" in regards to things like abortion. they just ignore the leap in logic because it's "permissible" by some protestants.

for many, Christianity is simply "turn the other cheek" on steroids, wrongfully interpreted as "mind my own business" and "only believe in the parts I want to".


neat! any support for efficient spatial queries?


Not yet! But that will be coming. We need to add it in for BitCraft.


Using Google assistant via a home mini and nest mini has generally gotten worse, not better, over the last few years. I can't wait to bin them and replace with something else that works well with home assistant and setting timers.


I have the same problem. I use Google home for voice interface to home assistant. I switched from Alexa a few years ago because Google had more open options for integration. But the Google home has gotten much slower in response time. I still have an Alexa in the garage and it is so much faster. I have a Google home speaker in most rooms and it has gotten worse and worse at knowing which speaker to respond with. Sometimes I will be standing right next to one in the office but the one in the kitchen will respond saying it couldn't understand (because it's too fricken far away).

I'm beginning to wonder if it is actually hardware/microphone degradation because I can't imagine the software degrading so far and so gradually.


spent so many hours playing this years ago.. great memories. too bad it doesn't seem to support inverting mouse y-axis or changing keybindings. too much muscle memory to overcome.


I’m pretty sure you can change your controls from the home page (not once you’re in game). Not sure if the settings have inverted y axis, but it had scroll to jump so I wouldn’t be surprised.


API for the server example looks... actually good, wow. Nice job!

Also tickled to see my erlang 1M comet blog post referenced. A lifetime ago now, pre-websockets.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: