Hi,

I have noticed of late more new members seem to post ChatGPT output to the forums. I have not personally used AI in the past so I thought it would be wise to check it out. I do a lot of command line work. So I presented known operations to Chatgpt to see what would be returned. I was amazed to the linearity of the output and close to what would be performed by me. Not that I'm sold on AI but somewhat surprised as to the depth and close to human interactions. If I do continue experimenting with AI then my actions will be based on my interpretation of the output.

I can see how a new user would/could fall into a trap with AI. Experience does matter along with in-depth understanding of the subject at hand along with common sense. Do not accept everything as solid or delivery of proper procedures. Be-aware!

+1

FWIW. My 2 cents:
I don't know if anyone has mentioned this, but...

It's not always possible to tell if a reply is AI-powered. I think it would be appreciated to have a note about the AI's use, as some users do (e.g., this is the reply I got from the AI). (Perhaps it could also be mentioned in the forum rules, I don't know.) Not as a rule, but out of politeness.

I've never used it either. I'm tempted to, but... if I start feeding the beast, it'll end up writing English as badly as I do. haha.

EDITED: I used google to translate/adjust my text.

I use Brave Search, there is an AI component that sometime is used for searches. I use it occasionally. I have posted results in another forum I am a member of (non technical), I always add this to the post "AI-generated answer. Please verify critical facts. " This what is shown at the bottom of the AI result by Brave AI.

I find it to be decent, but not always. It has come up with wrong answers, even to technical questions, also steered answers.

BTW, I do not consider this AI. It's just a search that puts together answers from various sites, the sites are referenced in the answer; a nice touch.

Hi,

I say this with tongue in cheek; Sadly AI is being integrated into our society daily in new forms. A lot a people feel AI is the answer to everything. I can see it can be of help in future research in the medical field as it matures. I do not think it is mature enough to use in daily engineering field tasks but will evolve into daily use by companies if not just because of economics. I would not want it designing an aircraft that I would be flying on but could be used in parallel with human design but AI does not have artificial general intelligence even if it could think like a human. Sentient is not ever going to be a AI trait.

I really do not think programmers can design common sense into AI or that AI itself could grow it self into using a human trait, which with not all humans having that trait.

Just my thoughts from a old Norwegian widower that is thankfully still learning.

+1 For your post and +100 for the words in quote!

As a person who deals with this technology every day. I will not call it artificial intelligence because intelligence in Greek, the ancient language of philosophers, is rooted in mind (nous). mind (Nous) can only have animate logical beings. So for me these are just machines, they have no soul or spirit whatsoever... So their logical distinction is not based on logical conclusions or concepts.
Anyway, back to our topic...

Whether we like it or not, these machines have now entered our lives, and the truth is that technically it is unknown where this situation will end.As I said, because we deal with these machines as a hobby every day, I know that things are moving at incredible speeds technologically, the evolution of machines is multifaceted and I have not included what we do not know ... The secrets of the 3-4 big ones who are fighting to see who will finish first...
So what should a Slacker know today who has no idea about this?

Use this table to better understand what the terminology of the new eras means...

• AI → The umbrella term covering all artificial intelligence.
  
• GAI / AGI → The ultimate (still hypothetical) goal: human-level or above general intelligence.
  
• Narrow AI → The current reality: specialized systems that perform one domain of tasks well.
  
• LLM → A type of Narrow AI focused on understanding and generating language.
  
• GPT → A specific LLM created by OpenAI, widely used worldwide.
  
• Agents → “Software robots” that use LLMs (like GPT) as their brain to autonomously achieve goals in the real world.

+-----------------------------+

• What is AI (Artificial Intelligence)

AI = Artificial Intelligence.
It is the general term for any system (hardware or software) that can perform tasks that usually require human intelligence.
These tasks can include:

• Voice and image recognition
  
• Natural language understanding
  
• Decision making
  
• Learning from data

�� AI example:

A program that recognizes whether a photo contains a cat or a dog.

A chatbot that gives answers based on predefined rules.

• What is GAI (General Artificial Intelligence)

GAI = General Artificial Intelligence (often called AGI = Artificial General Intelligence).
Here we are talking about a hypothetical type of AI that can:

• Understand and learn any mental task that humans can do
  
• Transfer knowledge from one field to another
  
• Has flexibility and creativity similar to humans

�� Example of GAI (hypothetical):
A robot that learns medicine, music, engineering and philosophy, and can apply them all without being trained separately from the beginning.

• What is LLM (Large Language Model)

LLM = Large Language Model.
It is a type of AI that specializes in understanding and producing natural language (text), trained on huge volumes of text.

LLM features:

• Based on neural networks (mainly transformer architecture)
  
• "Learns" the probabilities of words and phrases to produce coherent text
  
• Can answer questions, write texts, translate, summarize

�� LLM examples:

• GPT-4, GPT-5 (ChatGPT)
  
• Claude
  
• LLaMA
  
• Gemini

• What are Agents in AI

An AI Agent is an autonomous or semi-autonomous system that:

• Perceives the environment (via data, sensors, APIs, etc.)
  
• Thinks or makes decisions (according to a goal or programming)
  
• Acts to achieve its goal (e.g. sends emails, writes code, makes reservations)

�� Agents can have:

• LLM as a "brain" (e.g. GPT that decides next steps)
  
• Tools they use (e.g. access Google Search, databases)
  
• Memory (to remember previous actions/information)
  
• Autonomous action loop (e.g. continue working until a goal is achieved)

�� Examples of AI Agents:

• A bot that searches the internet, analyzes the results, and creates a report.
  
• A "personal assistant" that schedules appointments, sends emails, makes reservations.
  
• Agents in games that learn how to defeat opponents.

+-----------------------------+

How can you know if this message was written by me or a machine?

When you see in text/code :No need for further clarification on who wrote it...

There are already goodies for you to try at SBO. (lm-studio suggested)

Hi,

I do agree with your post and the AI generate output. Very linear and broad presentation. I have been trying AI over the last few weeks and find it useful to reinforce my procedures but not absolute. One of the reasons to try AI was to find out how close output would be to mine when doing simple procedures on my machines.

Never going to see a common AI general intelligence but just a way to worm things around the Internet as compared to human interventions or seeking relative information available with a search engine. One just needs to be specific with details when doing a search.

For me i see AI simply as: corporate piracy.

They took everything they could grab and ground/milled it thru the blender of "R&D" and serve it back as "intelligence" (I've been told that "I" stands for "intelligence" in the LLM acronym).

Besides obvious breach(-es?) of license that occurred, the scale and lame excuses ("Even if we don't do it the enemies will still do it") and several high profile scams ("A"ctual "I"ndians, fake bots etc...) I still have the ethic issue to use it.

It was stolen, the answer was "no" and they still took it - they once again show us their rapist mentality.

My answer is still "no", and they (bespoke AI leader companies) raped the whole internet. Again.

Just my 2c.

Hi,

I really do not like the use of rape of Internet fishing by bots or AI. If that is so then my use of a common based search engine would fall under that profile of rape. I've been using the Internet since it's inception and find the use has helped me through out my career and private life. So I cannot see using an available technology and turning back on it to garner information from the Internet be it AI or general searches as a problem.

If garnering information was a problem/rape then Google and many other corporations would be considered rapist too.

onebuck, don't take the word so literally. It was used in the nineties, especially in First Nations literature concerning dominant (canadian) capitalist/corporate practices. It really seemed to fit well regarding logging companies. When I was a young whipper snapper, we referred to strip-mining and mono-crop agriculture as rape. The common (vulgar) term for canola is "rapeseed" because of what it does to the landscape when let loose.

It was explained to me, however, that if any practice made money for the stockholders, it was "good business", and shouldn't we all get behind that?

Morality is always a relative stand.

Hi,

I take this situation very seriously and most people would say it's okay if it builds up the whole. Breaking the law or stating of same is not something that should be looked at lightly. Gleaning has been done for ages and it benefits the one who works to get something left behind be it a harvest of food stuff or even data across the vast Internet. For someone who say it is raping is stating something is wrong with that action when it's not wrong. If I'm wrong for taking someone's statement as literal then I am guilty.

Now if data that is copyrighted and used then that is a problem and government should take some action to regulate any actions of that sort. To use material that is protected by law should never be used without authorization by anyone.

No tongue and cheek here but someone who over the ages of my life find a lot of people do tend to bait or troll for reactions and sometimes I fall into the trap.

That might be appropriate but unfortunately it isn't true. The name comes from the Latin rapa meaning a turnip. Rape is actually a variety of turnip (Brassica rapa) cultivated for seed rather than for its root.

You might be amused to know that the Law of Moses (Deut 23:25) already makes clear the distinction between gleaning and wholesale theft:

"When you walk through your neighbour's cornfield, you may pick ears of corn and rub the husks off between your fingers, but you must not take a sickle to your neighbour's standing crop."

When I google for something (actually I use DDG, not Google), that is gleaning. It does no harm to any copyright holder. On the contrary: it helps copyright holders to get their work out to interested members of the public. But when the entire Internet is stuffed down the voracious throat of AI training, that is harvesting, not gleaning.

Hi,

I am sure that the algorithm for getting material across the Internet for AI has the means to differentiate copyright posted material. AI gleaning or worming across the Internet is no different than you doing a DDG search with specific terms with you deciding if the material is protected or not. Google or DDG has the means to search via a massed DB or storage to quickly provide their services but still depend on the user to differentiate materials presented as search results.

Your perspective is different than mine as to the definition of gleaning by us or AI rather than harvest. Pulling information from the Internet is no different for us but AI does have the means to classify data to be assimilated for AI that is presented to a user.

So by your definition we can pull information from the Internet but AI cannot do the same because it is formed Artificially.

Gentlemen, the entire world's knowledge is quantized.
If you can understand this, then you understand that the discussion is no longer meaningful.
The only thing that MAY make sense is to change the licenses in your projects to AGPLv3. In a world where most AI tools are cloud-based, AGPLv3 is the most protective and “good” for the community.
I followed a similar discussion on the SBO mailing list. I don't know if SBO allows a license change in SlackBuilds, but for those who have objections, it is the only license that IF respected at least benefits the community in some way.

What @onebuck says is the whole point: "So by your definition we can pull information from the Internet but AI cannot do the same because it is formed Artificially." (?)
And because this does not stand up to either law (license) or logic
but most importantly, in practice it cannot be controlled! Forget it.

PS. Don't be afraid of the machines, be afraid of the people behind them.

absolutely not! [c.f. OpenAI Says It's "Over" If It Can't Steal All Your Copyrighted Work]

Yes, AGPL is like GPL that it allows any user of any derivate work to get access to the complete source code including any build scripts used and to be free to rebuild the derivate work with or without any modifications. Compared to GPL, AGPL extends the view upon who is the user. For some kind of server software GPL considers the administrator of the server to be the user with the rights listed in the license terms. The AGPL license includes anyone connecting to the service from the network to also be a user with all the rights that the license gives.

I would consider a large language model partly trained on sources with an AGPL license to be a derivate work of that software. That would mean that any user connecting from a network to use that LLM would have the right to get access to all the training data and any scripts used to train the model to train his own model with or without any modifications to the training data or any scripts or settings.

regards Henrik