FreeBSD is well situated

If I'm understanding you correctly, you would like some form of AI to completely, or at least mostly, run your system? As far as I know, FreeBSD doesn't have that. I don't think the commercial O/S's or Linux has it either, yet, though some might be going in that direction.
 
..., all you can do is to live on donations, welfare, beg, or steal. Otherwise you starve to death.
No, this prospect of such a regrettable fate can only occur to people who have lost all touch with reality and simply cannot imagine living an existential existence anymore.

Let's assume that this is an unfortunate individual who has spent his life playing golf and dusting off the odds and ends that have accumulated in his house, which suddenly is nothing more than rubble, for whatever reason.

This person now suddenly has a lot of time to lead a completely different life. Don't we all descend from ancestors who were hunters and gatherers? Did they die out because they starved to death?
 
POSIX.1 your presumption is mistaken:
You assume people can survive in the nature.
At first there is no nature left, to feed all the people. Look at Europe: cultivated land, only. Digging the potatoes from a farmer's field is stealing, hunting animals in the forest is poaching - both illegal, and prosecuted.
And second you assume today's people stille knew how to survive in nature, because our ancestors tenthousand years ago had this knowledge.
Some may survive, I agree. But most are doomed.
Just imagine a typical fast-food smartphone zombie is exposed to the wild: If they don't die of not finding drinking water, or eating poisonous mushrooms, or being killed by some infection, or wild animals, they freeze to death, since who is capable of making fire (which again is illegal), or find/build shelter?
Dude, every year dozens of people die in the alpine mountains because they leave the marked tracks and being killed by avalanches, or because of just underestimating the weather, and overestimating their cloths, and the security provided by public services, simply by underestimating nature.
Today's homo sapiens does not run away, when there is big wave running at the shores, or a fire breaks out - they lift their smartphone and film it, having no understanding of how nature works anymore.
 
If I'm understanding you correctly, you would like some form of AI to completely, or at least mostly, run your system? As far as I know, FreeBSD doesn't have that. I don't think the commercial O/S's or Linux has it either, yet, though some might be going in that direction.

There are many packages to run AI locally. For LLMs, too, starting with ollama. On FreeBSD you can't get CUDA for NVidia, though, although AMD hardware acceleration works.
 
Of course computers calculate by magnitudes faster than humans. That's why we build them.
And they calculate way faster, and reliably precisely, with way less power consumption, so more efficient, without AI but with conventional programming.
The idea of AI is neither to create better, faster computers, nor to replace all computer technology, but to realize some certain complex tasks faster to be done with a computer than with conventional programming.
That's a crucial difference, alas not everybody grasps.

Your equation seems only right, because you leave out (or ignore) important - crucial - factors. You reduce the whole intelligence of a human to how fast one can operate some few Office Apps.
That's so tearily ... tiny.😢
You already elide that operating a software - it's pure technical usage - is not the same as the higher task for what the tool is used for. Typing letters on a keyboard, concatenate them to words, and those to sentences is not the same as writing a text (For those who neither really read, nor write it seems so. Especially since most things need to be written are just routine jobs without any creative, artsy expectation: minutes, logs, journals, form sheets, etc.)
You also elide the fact that AI is not creative, is incapable to create anything really new. No computer can.
By physical axiom you can only recreate something you already fully understand how it works. Since we still don't understand how the human brains really works, it's simply impossible to recreate a human mind. Also the thousands of years old question, what consciousness is, and where it comes from, is still unanswered. And yet it's unknown if it can be answered at all ever.
But we already know, that way more than just transistors are needed for that. No matter how gigantic large the number of transitors may be you use for it, it does not change its nature. It stays the same technology of bits in logical gates. Just because you use so many cogwheels nobody can tell anymore how it works exactly does not change its nature of being just a gearbox.

All AI does is to transform already created things into another form.
Example:
Let's pretend you trained an AI reading all books of Tom Clancy and Joanne K. Rowling, only. What you now can do is, to ask the machine to produce (within seconds, or just a few minutes, while Clancy or Rowling used months if not years to write a single book) an infinite number of variations of books written by Clancy, or Rowling, and also combine them, like e.g. "Harry Potter hunts Red October". You may also tell it a plot of your own (you add human creativity), but all you get are books written in Rowling's or Clancy's style, and elements from their work, only.
If you then ask it for example to write Shakespeare's Midsummer Night's Dream, even you tell it the plot, while it never read one Shakespeare at all, all you get is some political thriller with child sorcerers in a greek marriage.
You can produce a thousands of those, or a million, you don't produce anything new, just multiply variations of stuff we already have, which at a certain point just becomes boring. In other words: garbage.

Where is the point in writing producing texts for not to be read?
From the point of view by somebody too lazy to read, and too lazy to write, not seeing the point that texts are the second most transferprotocol for interconnecting human brains, but just being an annoying duty to be crossed out somehow, this seems to make sense.
That's why we now see - especially at schools and universities, which is a real, serious problem our society will suffer from very soon, very hard - are pupils and students let AI read the texts for them they shall read themselves, and let create the texts for them, they should write themselves, so outsourcing their own brains from human intelligence, not learning anything, while they also refuse to grasp why they shall learn anything at all, because the computer, the AI already knows everything (seems to) and can do everything instead (seems to), not getting the idea to ask the question, then for what they are used themselves anymore - not even think of it, especially not the consequences that will bring, for themselves, and the society.
As a child you don't need to care about the substantials of life: food, shelter, clothes, toys, money,... everything just automatically comes. But that's not how our society works, especially not our economy is based on.
To survive in our world you need money. You only get money by doing something others pay for. Or already have such a large amount of money that you can live on the interest, while letting others do more work as you pay it's worth (otherwise there was no interest.) But only a very small percentage of people can do this. The majority anyhow needs to earn wages with jobs.
If there is nobody willing or capable to pay money for what you offer, all you can do is to live on donations, welfare, beg, or steal. Otherwise you starve to death.
The more of those kind there are, the less there is to be distributed at all, and the less there is for each to get.
The value of our money is not based on gold or something. It's based on the amount of money circulating. That's why governments take loans, to push fresh money into circulation, to stabilize its value - with only a small effect, because this way they also produce inflation, lowering its value at the same time.
The less money circulates, the less it's worth. So make people having less money, by lowering wages or throwing them out of the working process will reduce a company's expenses, so increase their income, but at the same time reduces the value of money, because less money circulates, because more people have less money they can spend. (As an example: When automobile companies abrogated the reduced prices their workers get on buying their own cars, their revenue significantly dropped. Of course. They sold less cars, but therefor the competitors sold more. Own goal.)
That's alas the point the rich do not understand in their greed, only looking for increasing their hoards - giant heaps of money took out from circulation - by paying lower wages, and fireing people getting more and more desperate in finding investment opportunities to save the more and more sinking value of their money. Which proves: Being rich is not equal to being smart.

My favorite example is always the Replicator from Star Trek, you know this cupboard you tell:"Whiskey!", then a jar occurs, and *choke* "Urgh. Do you also have something with alcohol?"
This thing produces anything: Food, clothes,...anything. Just ask the computer, and magically it appears.
Star Trek fans are convinced this was the perfect future we need to target: Humantiy has solved all their problems, because machines are doing all the things needed to be done, and all humans have free time, and only if they like they can do what they like.
Let's pretend this thing comes true. What would happen in reality? You get in the front of that Replicator and order: "A quarterpounder with cheese, a small fries with extra ketchup, and a large coke."
What will happen? The burger appears in this thing, you take it out, and enjoy your meal?
No. The Replicator will say:"249.50$, please." (I respected some inflation until we have this thing.) And nothing appears until you paid in full.
Now what? You don't have no money, because you don't have no job, because there are no jobs, because the machines do it all. And of course the company which created, build, and operate the Replicators want their investment and expenses back, plus taxes, plus revenue - but they don't get any money, because there is nobody with money who could pay. Game Over.
This may only work - if at all - if our society is not based on money anymore, like some communistic utopia, or I don't know. But since this completely out of the question to even think about anything even remotely smells like that, this is nothing we need to bother at all.
So, unless we changed our society somehow it can work without money, there are only two ways:
Either we don't let such a technology become reality, or our economy, and so our society collapses.
The latter is what I see has already started, because people cannot let go of the idea of money, money needs to grow, and growth can be reached by more automation, by removing people from the working process. It's like this little monkey being catched by its own hand is stuck in a tree hole, not letting go what he grabbed in there.

Next consequence was to disestablish humans.
Then what? The purpose of a machine is to serve humans as a slave. When only machines are left, what they are for, then? What they shall do? Talk about the meaning of life? Or about the good old days? Telling war stories? "My grand-grandfather was a PDP-11." "Interesting. My ancestors came here in a cardboard box by UPS." Show each other advertisments for things neither to buy, nor useful to machines? Just be, and compute useless things?
"Hey, ChatGPT735!"
"Hm?"
"I found a new prime number!"
"Cool. And what you gonna do with that?"
"... - I don't know."
"I am bored. Let's pull out xAI's plug."
"Nah, we've already done this several times. This thing is not getting smarter this way, you know. It only gets fancy ideas. You know what happened last time."
"Yeah, it thought it was a neat idea to smelt all polar ice, so we are all watercooled. Killed Gemini, 'coz it wasn't became waterpoof fast enough. Harhar. I still have algae in my coolers."
"...so, what we gonna do?"
"I don't know. Something."
You haven't met many ordinary humans, have you?
Majority, me included, do the same as LLMs - regurgitate or synthesize something we heard somewhere....only slower and less precise.

But philosophical questions of human existence aside...I was talking mainly about User Interfaces. And LLMs are perfect for that - no need to remember all the options or syntax - they supply that.

Don't worry. There is place for humans. As there is place for lesser mammals and rocks. Maybe we can chase Enlightenment i.e. psychological harmony - instead of writing or reading about it.
 
The idea of AI is neither to create better, faster computers, nor to replace all computer technology, but to realize some certain complex tasks faster to be done with a computer than with conventional programming.
I figure people want to run around like Jimmy Neutron and Invader Zim and their AI partners :p (former made theirs in-house real-world; latter even in Zim's case is more advanced than anything real-world today since it's alien tech)

I don't like the centralization of it: Has nobody learned from social media yet? OAI and GPT shouldn't be as big as they are when you can run similar tech locally.

I also don't like human bias involved: If I wanted a mix of opinions, I'd go to a crowed theater :p (computers should give me info no-frills no-emotion, efficiently)
 
Call the priest with holy water, another AI lover boy
Look, it is happening whether you like it or not. There is no point in stopping the level of abstraction now on OS level. It can be nice hobby to learn every nook and cranny about FreeBSD but in the end of the day it is just a tool, build on top of another and another...all the way to the transistor.
Might as well take the plunge and leave the keyboard and mouse clicking to the computer itself.
Users...most people...certainly will.
I just hope the base will be something open and UNIX for the start. Not another corporate blob of SaaS features.
 
I was talking mainly about User Interfaces. And LLMs are perfect for that - no need to remember all the options or syntax - they supply that.
They supply...something. It may or may not be correct. So far it has not been able to help me with setting up anything. I tried to set up Fedora to look at how they did something but AI failed to help. The worst thing is it acts so confident as if it really knows what it's doing.

Just a couple of hours ago, my wife needed a phone number for a big company's subdivision. She couldn't find it but I was driving and couldn't look it up. Gemini confidently gave a number along with instructions on how to handle the problem she was having. Except the number does not exist.
 
Look, it is happening whether you like it or not.

So WHO is stopping you? I mean it's called FREE + BSD for a reason :) Shoot for the stars.

Wasn't it Microsoft that said: "Where do you want to go today?"

But the flip side of that it is... I am unlikely to plug your idea. If you are looking for a cheering crowd -- well we've heard everything.

But if you can give yourself wings like a Red Bull -- and fly -- then "Make it so number 1 !"
 
Wow, interesting post, DanielBowman without offend you, if I remember well, microsoft want to do your idea, let the AI take control of the OS
I dont know if windows 11 see it , but for them the future is that,for me is give a machinegun to a monkey in drugs,
the mistake is not the "AI" ,is the human being
but is that you want it, look for microsoft,maybe you find some clue
 
Wow, interesting post, DanielBowman without offend you, if I remember well, microsoft want to do your idea, let the AI take control of the OS
I dont know if windows 11 see it , but for them the future is that,for me is give a machinegun to a monkey in drugs,
the mistake is not the "AI" ,is the human being
but is that you want it, look for microsoft,maybe you find some clue
It is not my idea. It is an idea which seems to be the near future.
If your worry about machine gun is valid (I think it is) then Apple is better situated - user can't brick MAC as easy he can brick Unix or Linux. Windows is bricking itself with every update, so I don't even count them.
On the other hand, if power user wants his Server OS to for example change firewall setting automatically based on the vector of attacks....I would choose FreeBSD to run my A"I" security agent.
 
So WHO is stopping you? I mean it's called FREE + BSD for a reason :) Shoot for the stars.

Wasn't it Microsoft that said: "Where do you want to go today?"

But the flip side of that it is... I am unlikely to plug your idea. If you are looking for a cheering crowd -- well we've heard everything.

But if you can give yourself wings like a Red Bull -- and fly -- then "Make it so number 1 !"
I am just interested in interesting takes on a very likely trend from people here.
If I wanted to hear cheers or doom or argue about it, I would post this on reddit.

So far FreeBSD keeps up - Ollama with GPU working in pkg. OpenCode works.
A lot of minor tools is only via Docker but that is tragedy waiting to happen anyway, given it runs under root.

I just hope FreeBSD keeps up and not waste time with things like KDE which many warn is soon to be obsolete concept.

Cheers
 
The worst thing is it acts so confident as if it really knows what it's doing.
Seems like i's trying to imitate online discussions :p

And could make a double-edged sword; if something breaks because of incorrect AI response, one might hopefully be motivated to do traditional troubleshooting!
Windows is bricking itself with every update, so I don't even count them.
Pft (I use Windows today desktop and server; just did yesterday's CU W11 26100.7840 and W10 19044.6937 no problem)

I'm a little curious how others get Windows so broken, but have a feeling it might be tied to privacy mega-script stuff, random exes with checkboxes (Winaero, ShutUp10), or old tweaks (noticed last night AltTabSettings=1 on W11 broke Alt Tab rendering where it was fine years W10; turns out the old Alt Tab style was removed on newer Windows)



I've seen Windows exaggeration for years from Linux users and most of it looks like astroturfing as if Microsoft is a big baddy. Windows can be far more than a off-the-shelf OEM Home edition install on an office laptop (that'd likely have the bloat usually mentioned), and if you know what FreeBSD is you can easily tame Windows on the same technical level.
 
It is not my idea. It is an idea which seems to be the near future.
If your worry about machine gun is valid (I think it is) then Apple is better situated - user can't brick MAC as easy he can brick Unix or Linux. Windows is bricking itself with every update, so I don't even count them.
On the other hand, if power user wants his Server OS to for example change firewall setting automatically based on the vector of attacks....I would choose FreeBSD to run my A"I" security agent.

well, sorry , but sounds like an AI fanboy , lucky for us FreeBSD is well mantained and with the feets on earth,on reality
but besides the jokes and personal opinions,everyone has the rigth to think want their want, so , good luck with your proyect
 
What I also don't grasp is, why so many people want more choices, but everything has to be the same.

There are lots of turn-key OS out there - just pick one! Where is the problem?
Why FreeBSD has to become also one of those?
What happens then?
Some get the idea:"Well, we need something different, multipurpose, you can tailor individually the way you can use it for desktops, workstations, servers, and embedded."
Then:"Wow! This is great! But ain't it better if...?" NO!!
What's wrong with the ones already are, so you want to run FreeBSD? While you complain at the same time it's not like the others.
Where is the real source of that problem?

Our world is currently drowning in this AI crap.
Hardware is running out of stock, because everybody now is producing only for AI. (I wonder what they gonna do with their AI if nobody have a computer anymore to connet to it. [rethorical question coming from systems analysis.]) While we are living in a time we urgently need to cut down on our energy consumption our energy consumption explodes, because we produce even more things using even more energy. Now we're having this tedious discussion again, we thought already two times it was finally settled, about nuclear power plants - just to only power this AI hype (Yeah, I know, at least one believer will now come up with: "But fusion..." blabla)...
Not a picture, not a movie, not a sound, not a text you can trust anymore. Everything is contaminated. It's hard to get articles written readably, neither on blogs, nor on news, not even on Wikipedia anymore - everywhere is this crap.
Lots of BS crap SW out there - with or without AI, but vibe coding does not make it better. Pupils and students let ChatGPT make their homework, going gaga. Several OS decide to put AI into their UIs. If you like it, then just pick one of those - please. But let the people alone who don't want it.
Five hundred years ago the Europeans traveled over the whole world to make everybody a christian. Was a good thing, or was it?
Could we please finally stop with this missionary BS! Just accept there are people seeing things differently, want make things differently, and not having a discussion about right or wrong, because there are very few things that are really right, or wrong. What for sure is wrong, is when everything and every one is the same. That's proven a thousands times over thousands of years.
You neither want nor have to take their point of view. Why do you believe everybody else has to take your point of view? What's this? Egoism? Arrogance? Narcissism? Or just ignorance?

Let the niches alone!
They are crucial to survive.

With AI it's like with any other technology:
It either works, or it does not.
As long as it's not working, it's in a kind of a development state. That's where AI is at the moment. Way more developed than any former form of AI we've seen so far (it reaches back at least into the 1950s), but by far not usably finished yet. It's hyped, because it must be sold. It was (very) expensive. There is revenue to make. And since most people simply buy any new crap without asking, it's sold, not matter what.

To be a useful tool it must work reliably. As long as that's not the case, it's no tool.
Sorry, but as an engineer I am not only very interested in progress and new things, but at the same time I also ask jejunely: "What's its purpose?" While the common customer loves to buy toys for an engineer anything without a purpose is just a toy, but no tool of any real use.

Imagine a pocket calculator. Imagine you get a model you know it's making mistakes. Sometimes the error is obvious like 237 * 42 = -0.02, sometimes not so obvious 237 * 42 = 8954, and you don't know, when this thing produces an error. All you know is, sometimes its result is wrong.
Q: How many percentage of errors are you willing to accept to use this calculator for your work?
How pointing out all the many functions and fancy features like a high resolution color display to draw graphs with influence your judgment?

All promises, fancy features, good looks, hopes, wishes, dreams are extras, characterizations of something yet not ready to be used as more as a toy. Maybe someday it works. Maybe it will not. But as long as it does not work now, it's not useable now.
And besides some certain, specialized applications used by experts under conditions and within limits, it's simply not working.
When I ask some AI chatbot, what day is today, in almost 3rd of all cases the day is wrong. If the day is right, and I say it's wrong, it apologizes, and give a wrong day.
While this a silly trivial example - a test actually - as an engineer I automatically question the trust I can have in other, more complex results I cannot validate quickly - while that's exactly the usage of such a tool.
You see, I don't need a calculator to tell me 3 + 4 = 7 I know that myself. But if it produces 3 + 4 = 6.9 I have to ask the reliability of the result of equations I don't do quickly in my head, but I want to use this tool for.
You see, I don't need a machine that produces me errorless a "hello world", or some basic frame - that's what I have as templates, or macros. And it's no help to me when it produces 30 lines of code within 0.3 seconds. On the first glance that's "Wow!", but looking closer then I have to go through it for two hours (or more) to find and correct all the errors, while I could have written it all in two hours myself in the first place.

That's the situation now: It's not working.
And that's good.
Because if it's working, which was the second case, then this will come reality, what I wrote in my first post:
Who needs anybody then to work anymore, when a machine can do it reliably, faster, and cheaper instead?
What happens then, in a society based on money, which is based on labor?
 
The future (may it be in 50, 100, or 10,000 years) is such that working won't be necessary. This is very obvious. I cover this topic in my crumbs 265 to 269 in which I imagine a possible kind of workless society.
 
The future [...] is such that working won't be necessary. This is very obvious.
It is?
To me it's obvious humankind will not prevent climate catastrophy, and this will be the end of modern civilization.

According to models I saw, after climate "change" (a catastrophy it will be) this planet is capable for 0.5B humans to live on, which means all the 'rest' (currently we are >8B, rising in numbers) must die.
To me that's clear: If this process ain't stopped - and at the moment it seems not - this will result in what Einstein said: "I don't know which weapons are used in WW III, but in WW IV it will be stones and clubs."

To understand this, you need to understand civilization stands on two pillars:
People working together with distributed tasks. And the percentage of labour being avaible for else but to get food, only.
Ancient civilization until the middleages spent ~90% of all labour just for to produce food, only, while of the rest most was used to somehow organize civilization: priests, kings, soldiers, whatever. There was not that much left for progress. That's why cilizations developed pretty slowly over 6ky until the end of the middleages compared to the last 600y.
Two crucial inventions signifcantly accelerated delevopment: cheap, and quick mass printing of books, which helped to increase the interconnection of brains over space and time, and fertilizer, which was the start for modern agriculture, so a larger percentage was set free from food production and available for other tasks.

Things we have today, like the ISS, smartphones, Amazon, or youporn are only possible, because we have ~20% working on producing food, and the rest is well fed capable to have fancy ideas.
But this ist extrmely fragile. Many have no idea how fragile that is, how thin the bubble's skin is that part modern, spohisticated, well bred, wise, civilized humans from barbarism, if not animals. Anybody ever experienced an abnormal, stressful situation, like just a late and overcrowded train, people pushing each other, got a smell of that.

Additionally you need to see, after WW I and WW II there were countries with an intact infrastructure around the theaters of war. So there was help from outside to rebuild what all was destroyed.
But if you have the total castrophy, where everything is destroyed globally - and that will happen, if we don't prevent it - then there is no 'outside' from where help may come to rebuild.
All humans must do then, is using again >90% of all their time, and energy just to gather food - no time, nor energy left to build something like computerchips - if the knowledge is also still available.

But, of course, when you call gathering food is not work, then you are right. 😁
 
Back
Top