9:00 I'm surprised it actually gave an approximation of the right answer. LLMs are terrible at simple arithmetics, despite the massive matrix calculations they do in the backend...
@DistroTube I have a Question and a comment. Question: How long did it take for you to get really comfortable with talking into the mic? I tend to have a lot of dead air between my speech. I love the idea of the large language models. Depending on how you use it, it can be very useful :)
I would also add that "artificial intelligence" is simply a marketing term to make some evil corporation somewhere a lot more money. An engineer would consider the term "probability engine" to be more appropriate. If the engine has no data to work with, it can do nothing. The more data you feed it, the more it can look for patterns and their probabilities of occurrence, and therefore provide "the likeliest response" in any given scenario. But, by the same token, static data can become stagnant very quickly and sampling the wrong data can provide anomalous results - as has been shown with "glue on a pizza" for example. So please don't turn your brains off just yet, even though lazy younger generations today weren't taught to use their brains in the first place. Oh, and vim is better than emacs.
@@jackt-z2m I've really triggered you, sonny, haven't I? Nobody forces you to read my words. And in more than 30 years of working with UNIX and Linux, I've never once found a reason to use emacs - a lot of the time, I'd have needed to install it whereas vi or vim is usually "just there". If it works for you, go for it. Dry those tears and ask your parents why they let you grow up this way without saying the word "no" to you enough. That explains why the mere words of a stranger trigger you so badly. Discussion (three) closed.
Linux is for critical-thinking, non-lazy computer enthusiasts. AI is for lazy millennials that were never taught to think critically because of a poor education system and need something to do their thinking for them. Therefore the two are mutually exclusive in my universe.
@@alx8439 Thanks. But do understand that I am most definitely a boomer with decades of computing experience, both hobby and work, and I have managed to have a fun time doing both without any need for AI - so it's not something that I need today, just because it exists. To me, it will just remove a lot of the pleasure that I currently get messing around with computers.
@@terrydaktyllus1320 sure. I get the point. I treat LLM as a car (an automobile). You can get from point A to point B by walking, of by driving. Which might be faster in some circumstances, but not always. And when you just want to have a walk for joy - you can. Owning a car doesn't automatically introduce an obligation to drive it all the time
@@alx8439 But owning a car probably makes you (and me) a bit lazier because you'll use it to drive a half a mile to a local store as opposed to walking there. So your analogy proves my argument, not dispel it. Plus a professional athlete (who enjoys walking and exercise) will avoid a car as much as possible. By the same token a professional computer enthusiast like me avoids AI.
@@terrydaktyllus1320 absolutely. I'm not arguing. Just shared my own view. And I do agree, that youngsters tend to overuse AI instead of training their own brain - which is sad. On the good side I don't see it happening all that often now, given LLM models are not ideal. Yet
@@jackt-z2m Even the name "Artificial Intelligence" is a marketing term to make evil corporations more money. It is nothing more than a "probability engine". If you lack the engineering knowledge to see that for yourself, that is not the OP's problem. It is a scam.
@@jackt-z2m The fact that you refer to it by the marketing term "AI" and not the engineering term of "probability engine" shows you've fallen for the scam. There's no "intelligence" about it, it's just mathematics and statistics.
@@terrydaktyllus1320 K old man. I wish I wasn't so compelled to respond to this idiocy. AI can be used for more than just text generation. It can be used for small things, like generating commands to control a piece of software. Especially if you take time to fine-tune prompt or other form of input. That's just one example. Everybody knows you can't trust AI 100%. It's always a good idea to have fail-safes and other ways to prevent it breaking things. It can be very useful. Go to bed Grandpa.
@@jackt-z2m No, they probably just quite like working and being paid a salary - and doesn't want to replace themselves with a computer just to make a few billionaires a bit more money.
DUH! OBVIOUSLY! Should every one of the thousands who ALSO don't need/want it post a comment stating they don't need/want/value it? NARCISSISTIC, EGOCENTRIC INFАNTILE, grow up! Why'd you even open the video then, when it's obvious what it's about from the title? To whine?! _"A man has no right to occupy another man’s time unnecessarily."_ ~ John D. Rockefeller
It's amazing how software developed from 40 years ago is standing the test of time. I really love emacs.
It is always nice to see FOSS and A.I. morph together! Keep it up, I love this kind of content.
More to come!
Strange... I've always kept "FOSS" and "scam" very separate personally.
Why should A.I be a scam?@@terrydaktyllus1320
TNX,DT!!! Informative and concise!
Woooo nice! BTW, which WM are you using currently?
qtile. He ditched xmonad for now
Massively Cool.
If you haven't already, you should try different models from huggingface. You can write a Modelfile and have ollama use the model that way.
Clearly the scammers are catching on now - start with one "AI" scam and then lots of others want a piece of the gullible.
oh , this is nice
One day a few years ago, DT tried eMacs and he never made it back. Be warned.
I think you mixed up the key bindings for grammar and wording.
Yep, it's fixed in the config in my dotfiles repo.
Shouldn't "AI" have picked that up?
9:00 I'm surprised it actually gave an approximation of the right answer. LLMs are terrible at simple arithmetics, despite the massive matrix calculations they do in the backend...
@DistroTube I have a Question and a comment.
Question: How long did it take for you to get really comfortable with talking into the mic? I tend to have a lot of dead air between my speech.
I love the idea of the large language models. Depending on how you use it, it can be very useful :)
Idk if you notice...but he edits it very well. There are veryyy slight jump cuts though.
Hey DT! What would be an easy way to share a specific bspwm window to another computer? I've been trying it with VNC but it's not working... :/
Real tasks, where LLM models shine with programming tasks is code autocompletion
Looks interesting,, I failed loading LLAMA 3.1 and minstrial
Make sure you pull them in with ollama. ollama pull llama3.1
@@DistroTube Thank you very much.
I would still pay good money to see DT play Command and Conquer Tiberian Dawn NOD Campaign. OpenRA is a thing...
Maybe it didn't use GPU
least but not last
no thanks ...
last
Congrats
"the last will be the first" (Good sort its paradise existences with LIFO)
Are the rumors true you sold DTOS to Amazon?
Sad if true.
I would also add that "artificial intelligence" is simply a marketing term to make some evil corporation somewhere a lot more money.
An engineer would consider the term "probability engine" to be more appropriate. If the engine has no data to work with, it can do nothing. The more data you feed it, the more it can look for patterns and their probabilities of occurrence, and therefore provide "the likeliest response" in any given scenario.
But, by the same token, static data can become stagnant very quickly and sampling the wrong data can provide anomalous results - as has been shown with "glue on a pizza" for example.
So please don't turn your brains off just yet, even though lazy younger generations today weren't taught to use their brains in the first place.
Oh, and vim is better than emacs.
Emacs is king of editors. Grandpa should C-x C-c himself.
@@jackt-z2m I've really triggered you, sonny, haven't I?
Nobody forces you to read my words.
And in more than 30 years of working with UNIX and Linux, I've never once found a reason to use emacs - a lot of the time, I'd have needed to install it whereas vi or vim is usually "just there".
If it works for you, go for it. Dry those tears and ask your parents why they let you grow up this way without saying the word "no" to you enough. That explains why the mere words of a stranger trigger you so badly.
Discussion (three) closed.
Linux is for critical-thinking, non-lazy computer enthusiasts.
AI is for lazy millennials that were never taught to think critically because of a poor education system and need something to do their thinking for them.
Therefore the two are mutually exclusive in my universe.
Haha. I love the way you put it together!
@@alx8439 Thanks. But do understand that I am most definitely a boomer with decades of computing experience, both hobby and work, and I have managed to have a fun time doing both without any need for AI - so it's not something that I need today, just because it exists. To me, it will just remove a lot of the pleasure that I currently get messing around with computers.
@@terrydaktyllus1320 sure. I get the point. I treat LLM as a car (an automobile). You can get from point A to point B by walking, of by driving. Which might be faster in some circumstances, but not always. And when you just want to have a walk for joy - you can. Owning a car doesn't automatically introduce an obligation to drive it all the time
@@alx8439 But owning a car probably makes you (and me) a bit lazier because you'll use it to drive a half a mile to a local store as opposed to walking there.
So your analogy proves my argument, not dispel it.
Plus a professional athlete (who enjoys walking and exercise) will avoid a car as much as possible.
By the same token a professional computer enthusiast like me avoids AI.
@@terrydaktyllus1320 absolutely. I'm not arguing. Just shared my own view. And I do agree, that youngsters tend to overuse AI instead of training their own brain - which is sad. On the good side I don't see it happening all that often now, given LLM models are not ideal. Yet
No thanks. I don't want any of that scammy BS in my Linux.
You, sir, are so very ignorant. There is nothing inherently bad or "scammy" about AI.
@@jackt-z2m lol OK. Yes, ChatGPT is totally gonna change the world by telling women to smoke cigarettes during their pregnancies. Sure thing.
@@jackt-z2m Even the name "Artificial Intelligence" is a marketing term to make evil corporations more money. It is nothing more than a "probability engine". If you lack the engineering knowledge to see that for yourself, that is not the OP's problem.
It is a scam.
@@jackt-z2m The fact that you refer to it by the marketing term "AI" and not the engineering term of "probability engine" shows you've fallen for the scam. There's no "intelligence" about it, it's just mathematics and statistics.
@@terrydaktyllus1320 K old man. I wish I wasn't so compelled to respond to this idiocy. AI can be used for more than just text generation. It can be used for small things, like generating commands to control a piece of software. Especially if you take time to fine-tune prompt or other form of input. That's just one example. Everybody knows you can't trust AI 100%. It's always a good idea to have fail-safes and other ways to prevent it breaking things. It can be very useful.
Go to bed Grandpa.
0:10 absolutely fantastic piece of sh..
I have ip port spofing, or something like that, from start installing a models to ollama.....
Well, for those who need it... I won't.
Negative Nancy
@@jackt-z2m No, they probably just quite like working and being paid a salary - and doesn't want to replace themselves with a computer just to make a few billionaires a bit more money.
@@jackt-z2m Scambaited Suzie.
DUH! OBVIOUSLY!
Should every one of the thousands who ALSO don't need/want it post a comment stating they don't need/want/value it?
NARCISSISTIC, EGOCENTRIC INFАNTILE, grow up!
Why'd you even open the video then, when it's obvious what it's about from the title? To whine?!
_"A man has no right to occupy another man’s time unnecessarily."_ ~ John D. Rockefeller