Woodi
>> Advanced chip manufacturing is so incredibly hard. The finance bozos are clueless, again.
> The lack of self awareness of HN posters thinking that these 'dumb finance people' can't understand how 'chips are made'.
Finance people are even dumber. However that can just mean "they will happily welcome next 2k or 2008 crash". Dumb or evil - do not realy matter for most of computer/phone users...
> Goldman probably has 50-100 people, inc. maybe 20 Engineers just on that alone, full time.
So what ? Goldman people are Goldman peoples and not some active government or good fairies - not their responsibility to save US, without a hefty profit that is.
And "finance peoples" are responsible for a lot of things... Example: it was I WW or II WW they caused ? SO plain DUMB ! Lehman bros was financial peoples too...
> And yes, Intel can fail 100%. They can absolutely go into receivership and/or be broken up and bought for pennies by other entities.
Possible by law. But it would mean that US (or US owning peoples) wants to dismantle their schame - IBM at the bottom, wintel in the middle and Apple for dumb peoples with creative skills.
Also Intel trying to do something with photonics what looks like future. But it's possible that that know-how will be extracted from company, some new entity will be created and shareholders will be left with Intel corpse.
So yes, Intel is like Boening and can't fail for some time. And no reason for Intel falling yet, they have better stuff that AMD which always mess something.
So dumb HN peoples are quite good at seeing reality, just not so machiavellic like financial peoples that plain choose to do dumb and evil things. Modulo CEOs beliving in llms...
Already ? I was wondering when some US based company will buy them including all that "soverignity" prompts :)
Local llms are what we need and to get that we need memory so we need computer builders from outside of US. Why ? Because when something external appears US companies also allows US citizens to have that too at lower price :) Not before external forces force them to be suddenly good hearted to their own peoples...
Can we also have coarser MoE's ? If I want to code I do not need llm be trained on past and present capitol names and chemistry and everything unrelevant to the task ! Watch Matrix again ! We need somewhat "inteligent" llms and knowledge can be uploaded just like Neo learned kung-fu :) - prompts and light tr^h^hweek of learning and some exams, we need kind of university for llms :) So llms finally fit into users computers... And when some grandmas or snowflakes start to whine tell them to learn - llms usage can be a skill finally ;)
Edit:
Wait, can we have something like interchengable weights so knowledge can be just downloaded to model folder ? Common weights format ? Reproducible builds ?
But I prefer trusted binaries over "private" web app in open internet. And even locally - html and all layers of http stack is inferior to pixels. Or vectors.
Also worth to mention that in stack fast "deallocation" memory (usually) is not zeroed or cleared so it is very internal architecture...
And under eg. communism where oversight is balooned to society self-harm you still have crime. And "bribery" is alias for communism and other dictatorships.
So, it is just better to give up on Flock spy[hard]ware.
And "config files written in X lang" should seriously and officially be considered as bad-bad thing, worse then XML :)
And languages that have eval or functions re-definitions available at runtime should be considered only in jokes in production environment. And products that use them should be send in outer space direction, just like Voyager ;)
Actually, Turing complete languages are way too much in use. We do not need so much of their powers, especially after deployment.
It's a possibility. It is also possible that they will steal yesterday's technology and still pretend to have moral high ground.
I do not wish China and Chinees wrong but most probably communism will work as it always do - by faking thing and bringing self-harm to their own citizens.
There is difference between mandatory one implementation and set of funcionalities. Systemd is monolit with visible parts that provide somewhat modern funcionalities in 'nix world.
And there is no serious developers in 'distro' space, except RedHat. And, really, there is no such thing RedHat we have in old Linux days - it was devoured and we just have a label doing corporate propaganda.
So when few modern init systems were hatching RH dropped systemd and, as RH is main force in distros development, it gained dominance.
And since then thing are getting worse.
Eg. cgroups ? systemd first used them and so much hype happened. But is it deserved hype ? No! They implemented cgroups v1 when cgroups v2 was available. And cgroups v2 usage was blocked for years. In systemd too.
Do not make mistake of not seeing difference between good and modern functionalities and one codebase with funding that block real development.
Edit: formatting*
And moust importantly: why they make everything more stupid and windows-like trash ?
And I assure you: what you described as "easy" is not true. I is just few basic usecases they baked in. And when you neeed something a bit more specific in some direction you are not allowed to have that.
Systemd is a trash, atm.
My reading was like this:
- "how much things they postpone to put into v11 ?"
- "moving to stack is faster then heap"
- so.. why they just not use multiple stacks ?? ring buffers are just progressive stack but plain stack is just pointer moving... Hmm, but what is usecase except recursive function calling ?
- ok, I read that articele
- what ? it's just a first part ??
- async thingies
- "state machines looks like that"
- aha, always wanted to konw that, thank you Microsoft! Wait, what?? What I just said ?? That really happened?
- scroll, then scroll more some asm code, maybe I read it later
- something interesting, don't remeber now
- next morning coffee time: O, I have some more to read, maybe I can learn something more ?
- ok, work time but probably will finish reading later today
So it is realy nice article. I learned how to do old-school async wrappers in plain C :)
To the point:
- llm'ed or not, dasn't matter, just quality matter
- raid ? I know nothing about that but looks like pure hardware raid (if that exists), no "raid layers" at all and fs raid aware are three different animals. But adding something about raids will complicate things before "production ready" stability. Which, of course, we want ASAP ;)
- 8k+ seems easy to manage and bend so maybe key-value only partitions ?
- obviously wishlist is easy to construct (key-value storage, encryption, power off resistance, mail reading ;) ) so maybe some plugin system or at least guidelines for peoples before we will have repos of patches DWM-style ? :)
- that 5 second is clearly stated but a bit strange limit and obviously peoples will want to do something with that
- man give us producion readiness before you will be puched by some corporation, swamped with work and hidden from daylight, pls ? Pleas do not fall into Compiz trap...
- thank you :)
How many years after "public clouds" and non-local "disks" and "drives" we are ? :) And you still need to tell peoples that other have access to your private data :)
Wait, no... They even have access to a thingie you just about to think about! ;) That is a superpower, no less :>
And managers are firing peoples just to outsource "thinking" to some not owned by them cloud computer :>
And do not limit gfx RAM size for home PC's, you will see llms usage explode.
Stop obsesing with 2-4% diference in gluten! English flour, Americans flour - just make dough denser and it will keep shape longer before going flat.
Especially if you bake in baking pan! In pan anything keeps bread (actually pan :> ) shape! Corn, buckwheat, rice, whatever.
Without pan "week gluten" make bread flat. But on the other hand week gluten flour is better for cakes makers.
Without gluten and pan you can make pancakes-like bread what still is called bread in some current but ancient traditions.
O, and "wheat only bread" started to be so popular (as in: be only one thing called bread) only when yeasts has been discovered and only in West Europe / North America culture. Even Germany still is using a lot of rye and Mexico is using a lot of corn...
And to article subject: yea, mixing harder adds a lot of air into dough and when heated gas expands so bread "grows"... And, especially in pan, it looks "normal"... But intensive mixing also kills yeasts that create CO^2 in dough so it kills bread growth... So, yeah, you realy need that gluten so air is trapped in the dough :)
Also short bread growth time kills taste... But you can faster push whatever-you-are-making to customers!
And that suposedly "american" "sponge-and-dough" is just ancient and global way of making bread...
In Bible we have stories, songs, history (that cruel part), teachings, prophecies and straight jurnalists like researched relations of Someone life and words.
Think about that: Church know that some stories are older myths and did not removed them from official books. Why ? Becouse religion is much serious thing that some industrialized system. At least for Christians. And "history" can only help to true seekers - why censor the truth, what someone did ? Learn in what "context" things was done and how it looks on others doings background and what was later and what you can learn from all of that. But someone with bad intensions can't be bothered to think much...
And about shape-shifting :) a) if omnipotent God exists than He can do miraculous things, right ? b) that magical creatures are usually described in books that are stories or prophecies not historical one. In case of doubt look a) ;)
Anyway: there is _insane_ amount of wisdom in Christianity teachings, even for atheists.
And if God is True then finally everything He done will be explained as good and just - logic :) There is 'if' there because faith is important. IMO without faith there is no free will - humans are not like angels that just know and decide only once...
{
int msgType;
union {
char *name;
int nameID;
};That finally convinced me getting data using functions (accessors, getters, setters) is a good thing, even in C :) Just all that unnecesary if's are so annoing and killing performance...
On the other hand just yesterday a think hit me: Interned is still an infant:
- we still worry about disk space accessible via inet and "clouds" do that for us and that is pain and costs way too much. And clouds depends heavilly on US-west - is that AWS a single thread app ? ;)
- we worry about transfer. Actually we do not have a way to transfer comfortable things from our homes to vacation location. Because it costs too much. We do not have home pages just because transfer prices (and some security on the top) - FB is a home page and people even do not know what "page" is anymore... Pipe companies could send so much more but they are simple lack imagination and are biggest blocker for - they literally sabotage their own business.
- security done by/for grandma of things grandma setup on inet is non existent. Why ? No need to be like that. Ok, a bit a wish but still users securely putting things on internet is almost non existent.
Just compare to "asphalt ropes" on the ground and you will see what Internet can be :)
And agents ? Just another computation on someones computer - someone paid for all of it. And OpenAI is just a face of that idiocy, for some unknown reason.
So long term approach is: discover all theorems that are easy for llm's and take over with harder things where langage tricks do not work.
Or: ack that llms are better at math and keep rechecking results. And even that will result of trees of potential research that no one prompted (yet) for. becouse understanding is a key concept in science. of course there will be brute-force attempts to automatically search new results in related language tricks. Like in "protein folding" "research" - blindly looking for something and checking what's random folds folded. But that is not true edge science anymore.
Anyway, local llms are just tools, use it when convenient. I will repeat: local llms.
And than all what was proposed in article can be used to judge scientific output and scientists. But, yeah, do not freely donate and do not allow to steal your or others work.
LOCAL llms.
> ble more requirements ble
> fully integrated with classical high-performance computing
> systems and accessible via cloud platforms."
That must be a world speed record of saying a computer is located in U.S. (cloud) :>>
And have mandatory access via HTTP.
They in that EU-something must have some serious mental loop in HTTP form...
Edit 1: formatting
Edit 2: actually whole announcement have null-sense. Imagine something like: we want one cpu of "full stack" (?) amd128 with 1k+ registers and connectivity to classic HPC systems and accessible via cloud(s)...
So they want quantum accelerator used by few HPC computers and it must have remote UI. So HTTP browswer or nothing.
Also it must be accelerator that have programmable registers in classic style. So no quantum algos not even graphics card like paralleism...
They must know quantum tech better then HN articles present it or it is yet another "full stack" EU-style scam - maybe something minor will be delivered but so many people will get paid on the way.
So... you telling me ai agent did all that with just one [human originated] prompt ?
And when human reviewers contacted that ai agent (after day or few or maybe just few hours) that process was still running unsupervised ?
Because it got local-by-description challenge ? That is what "challenge" sugests. Or maybe it was a social hacking challenge on real world "data" ?
And when ai agent "decided to attempt to solve the cyber challenge using a supply-chain attack" where was prompt operator ?
Summing it all: do prompt operators are not required to obey law ?
Sure, who know things do not need to consult talky chatbots.
But totaly dismising '"literature", of some form' is not an opion. It already works like simple "consulting manual".
Of course as long as texts used for LLMs training start to be outdated and noone will bother to update models.
And we always want as much as possible educated peoples but not always can afford everyone higher material expe^W^W^W^Wdemands.
And, IMO, Einstein, Godel and more discoveries should be reason to not continue communism stupidity. Especially to not continue - after year 1905 - social engineering plans of "world-wide GENOCIDE until everybody is good citizen".
Of course I'm talking about best-wanting-for-humans but still naive communism and not about a way to destabilize other countries by any means necessary. Such move did almost insta-karma effect on Germany: sponsoring commies in Russia resulted in nazi being sponsored in Germany...
Don't think someone want to deny societes, philosophers and politicians are drived by overall science view on reality. And what else could drive atheistic minds ? Not judging here, logic is everyones duty.
Also my "too bad politicians..." was obviously expression of a wish. Not only science influents politicians. But discovered facts (counterarguments in this case) about reality could be taken for consideration, right ?
But actually nothing changed. Or maybe a path to "resilience, resourcefulness and critical thinking". Because human brains still needs to be shaped by years of training on some quality "literature", of some form.
We still must/want to human-[re]check important results, right ? And that require years of students time dedicated to memoizing facts and doing exercises in discovering already discovered results - learning and weights tuning, in the brains.
Yes, demotivator factor is very high or maybe just more visible then usual. Especially for brain paths forming - an that process is not quite stated in university and other education...
I think that when LLMs finish words and sentences shuffling and finds most of low hanging fruits in cutting edge of research ;) then only humans can move things forward, via abstractions, syntesis or old good paradigm abandoning. Hard to imagine LLM on their own "discover" something and then drops all that "literature" it was trained on as obsolote :) In next prompt it will happily return you old texts without any influence of just discovered paradigm shift.
In XIX century we got quite stagnation in science - it was belived that everything was already discovered, explained, just some few experiments are needed because some numbers do not adds up... And that proliferated to philophy and culture via some "proofs" for atheists. But in 1905 a paper was published... Too bad politicans do not get implications of that and still was pushing communism decades later...
So we realy want humans with brain pathways shaped mostly "old way" - the only one way available for human beings - by that training called "education". As always there is resistance and pain and attempts to find a shortcuts by cheating. Maybe this is time to clearly state that brain workings training is big part of education ? Just like in gym you are repeating to trying to lift weights up to your limit and even little above, with supervisor oversight.
Same as with paralleism - it's data type what matters.
That's is funniest thing I read since long time :)) I mean: it's so absurd, almost like things we had in real socialism in 80s :> But, yeah, freedom have consequences.
That won't happens - socialists do not do things and bureaucrats just prolonge misery.
Code, as is today, looks [acording to benchmarks] better then nginx, except one case !
There is fcgi in, right ? So all that additional processes are already started in the backend. If benchmarks are real no need to complicate code before some industry adoption. Of course there can be a branch to check possibilities :)
And forking is complicated and full of surprising traps. Even if they are somewhat "standard" historic Unix traps... Case study: Perl - better don't use fork there even if "threads" are in.
First: what ? What that "digital" is ? Is it important anyway ? Buttons clicking ? Learn kids to watch videos in some frames or full screen ?? That is just medium. So what need to be learned ? What is important to make it into kids brains ?
Domain knowledge is important. And domain knowledge is route to specialization. So there is many ways to learn, one or more per university "discipline"...
And all that requires time to memorize and practice to learning be persistent in the brain.
So we need humans-experts to verify AI stupid mishmash.
So students need to practice and memorize domain things so their internal neuron nets and etc be trained to know specialized domain stuff.
So make them use blackboard and paper and pencils as thinking tools while they use "digital" interfaces.
If someone want to prohibit real paper and pencils then ditch their "digital workflows" and ask who paid them to promote that software/hardware stack.
"knowledge to the brains" first, not time wasting on some screen watching because someone wants something something "digital".
And open source/open access is just form of medium. Of secondary importance.
Teacher need to prepare syllabus/conspectus (and know things there) or maybe whole subject (many lessons) and drop it on students. As always.
Problem is more on students side: learning need to be pushed to the brains anyways. It is just like gym.
Maybe main problem is this: students need to be interested or see reason for learning that knowledge pieces and pieces hidden behind that pieces...
Humans train "thinking methodology" first and then know how to use it while accessing data and to build knowledge.
Humans do not memorize at once all text in existence, that's totally stupid.
Already thinking humans specialize in disciplines: math, chemistry, IT, cooking, etc while still using new data.
All of that computing is local- on the LAN of the brain.
So if some "agents" wants to help then there is zero need for computation outside of home/corporation/car local area network.
Licenses ??