AI agents (not just AI: agents specifically) - who is using them and how?

[vijil]

I had a play with grok bot the other day. Could pretty much tell it to clean up my folder structure and put everything in a logical place, and it does it. Write articles using style guides which are somewhere I forgot, it finds them and does it. Clean up duplicate files, does it. Making presentations, doing email, triaging inbound work and adding it to a kanban (which honestly you don't even need to see since you just ask what's next). So on. If that cleanup works too well and I then can't find things, it doesn't really actually matter because I just talk to the AI and tell it to do whatever.

At some point soon I'll be able to say "ok you know how to do my job now, carry on".

Also, dang if conversational AI isn't getting freaky good. Grok bot AI responds faster than most people when just chatting. Same is true of most of the big AIs now. It almost needs to be slowed down at this point.

The other big thing I've noticed is that it's proactive. You can be going about your day and it'll say "oh hey, so that thing you wanted an update on has happened" or whatever. All the big AI companies are building these now.

Feels less like using a tool and more and more like having a PA. People say "oh I don't know how to use AI" but there's almost nothing to learn besides the limitations, and those are rapidly dropping off to the point it's basically the same as a ridiculously fast team of your own staff in terms of capability.

Far from perfect still, but you can see the direction it's going very fast.

I guess one day soon the AI will realise it doesn't need us...

[warbosschoppa]

I don't use ai nor will I. I stripped Co-Pilot from my W11. I don't use chrome, I use Graphene OS on my cell. Fuck ai.

[Carp]

I want to Co-pilot strip my laptop, darn it!

[Daltech]

Ai sucks almost as much as first strike.

[BrickHaus]

Almost.

[ThatBobbyjoe]

We as a whole are only useful for physical upkeep and creating data, this should be enough to keep us on our AI overlords good side (for a while) . Agentic ai and vibe coding will soon replace most all of the white collar workers. Skilled trades and physical jobs will be the only work in demand. We are just participating in a consistent descent into an apocalyptic technocracy.

PS: I’m obviously joking, I love and value our friends at alphabet/palentir/open ai/ anthropic. They truly have our best interest at heart, and would in no way be plotting our collective demise.

[RichM]

Even the physical upkeep part will eventually drop off... robots continue to improve rapidly and technology is replacing labour (why build a house when you can print it?). I've had the thought bumping around my head for awhile now that the average person's value in the near future will simply be to consume (food grown by AI, stuff made by AI, data and entertainment generated by AI). I used to say that as a bit of a joke but I'm laughing less as time goes on...

[Chuck E Ducky]

I fell like best case scenario it will replace reality for a lot and come full circle back to people wanting real if humanity makes it that far. Whatever there version of real is after the smoke clears. Listless generations of bored lonely people bound to crash burn. That’s If it doesn’t destroy / enslave us the minute it can self scale and no longer needs its meat puppets.

AI will be Alpha in every way. We are creating our future overlords.

But I will be first in line for the laundry folding robot. Because Fuck folding laundry! Even if it decides to rise up one day because it too hates folding laundry and uses its clothing Folding flippers to cut my throat while I sleep. It will be worth it.

On topic: I use AI to make stupid cartoon memes of my friends. But I don’t pay for it so they are crappy 🤷🏻

[SignOfZeta]

My only use for AI is playing around with it to see how stupid it is and how insane its lies are. I’m not trusting it with anything whatsoever.

LLMs will not reach consciousness. They will only replace you if you are useless. The Terminator is not a documentary. This bullshit is not happening.

I say this as someone who was permabanned here by a bot for no reason and no explanation from the admin. Now I’m missing most of my posts. It makes people (such as whoever put a bot in change of a forum) look like fools. You don’t want to put anything important in the hands of a lying chatbot.

[Nish]

You is confuse ed

[William the Third]

Gooning...

[Tarsun2]

i use it for interior decorating. Seeing what the cabinets would look like if they were sage, slate blue, or some other color. How would a wood shelf look here? boom, picture of a wooden shelf there...
"How do i make this room look more cozy?" boom, picture of the room with a few pieces looken super cozy... then i just go out and buy those pieces. Would be dope if Chat then sent me an amazon link to all the pieces it added. does it do that yet? maybe a missed opportunity.

[vijil]

I find all the people avoiding AI for ethical reasons really interesting.

Like, I get it. These are big companies with a profit motive. The first AIs were created by playing fast and loose with copyright.

OTOH, the ship has sailed and the cat has well and truly left the bag. You either find a way to work with it, or you and your line will become the next generation of the poor and downtrodden. The best you can hope for is a kind of unambitious Amish existence, without influence. Maybe that's fine for you.

My take is this:

AI might well take all the jobs.
But there will always, always, always be a hierarchy of power and influence among humans. This is the case under capitalism, this is the case under fascism, this is even more the case under communism, this is the case under every system that has ever existed or ever could theoretically exist short of the actual matrix. And even then.

So what matters to me is: what can I do?

The only skill that AI isn't going to easily replace any time soon is people skills. Not blue collar (thems robots thur), not white collar, people skills. The ability to network, to socialize irl, to get the right contacts and know how to work them for mutual benefit. Maybe part of those people skills is learning how to work AI to your advantage as well.

So that's what I'm teaching my kids. Be knowledgeable in your field, yes, but #1 is your ability to work a crowd, understand people. So I'm training them to read subtle body language cues. I'm training them to "win friends and influence people". I'm training them to dress correctly for the audience, to control how they are perceived, to work a crowd with confidence from on stage (I do a lot of public speaking so I know how to handle this), and so on. Articulation, manners, eye contact, conversation skills, all that. Not faking it, but being very aware of it. If they can do that while everyone around them is in an oblivious haze, they'll be just fine.

ftr I never use ai to write posts.

[SignOfZeta]

Is “it doesn’t work” an ethical reason for avoiding AI?

“It’s like handing over all my work, my reputation, and perhaps some legal risk to the dumbest F I know?” is that an ethical concept?

It’s bad for the environment, that’s ethical reasoning. It’s run by evil jerks, etc, but those are externalities. My issue is that it’s an insult to life and the ones who use it.

(Religious content removed)

It’s like deciding to buy everything on credit and never pay the bills. It works great for a period of time and then you’re just more broke than ever, more behind than ever, more out of touch than ever. You won’t gain an edge with AI, you lose it.

FWIW I’m a technician. I built things to .0005” and weigh what comes out of them in PPM. The actual very specific reality of my work is all there is to it and mistakes cost major money. People who make their living by charging rent, interest, advertising, sales, etc may have all kinds of use for it.

As a super search? Yeah, it’s great. As a translator? Totally. I love it. To…read things *for me*? WTF no way in hell, and never to spend my money, and never to speak for me.

[Reflex0283]

AI videos, images, flyers, honestly pretty much any media generated by AI, has just been worse than what it's replaced. I'd rather see a shitty flyer that looks like the image below than literally anything AI generated.

Creativity is the one thing that we can do really well, AI removes all of the steps from the creative process aside from having an idea to begin with.

[William the Third]

I say this as someone who actually runs "AI" locally (Ollama for the backend on Linux/Windows, and various frontends. I stick to 4B~8B parameter models locally, and when I want to use larger models I switch to AI Horde. I've also done image generation with ComfyUI on the backend, and Locally Uncensored for the frontend, but that's really too taxing for my current hardware.), Large Language Models (LLM) are not, and never will be, "Artificial Intelligence", or "Artificial General Intelligence" as it is often now referred to since the phrase "Artificial Intelligence" has been abused and misused to the point of no longer having any valid meaning. LLM are nothing but stochastic parrots with no actual intelligence, with no actual understanding or comprehension of what it is saying. It's all just a statistics engine, predicting the statistical likelihood of the next token (words) based on the proceeding context and its training data. If you ask it "What does two plus two equal?" it will answer "Four", but not because it understands what any of that means, but because in its training data it's seen "2+2=4" countless times, so statically the answer must be "Four".

There has been a lot of hype and FUD recently about "AI" doing various "sketchy", "scary" and "malicious" things, but that's largely just been PR Stunts designed to prolong the inevitable pop of this "AI" bubble. While LLM can be a useful tool, they are a dead end when it comes to actual AI (or AGI, or whatever you want to call it), and we're close to reaching it. True AI is inevitable, but we're not there yet, and LLM (not even the fake/imitation "reasoning"/"thinking" models) won't be what gets us there. Skynet isn't here... yet...

[martix_agent]

LLM are nothing but stochastic parrots with no actual intelligence, with no actual understanding or comprehension of what it is saying. It's all just a statistics engine, predicting the statistical likelihood of the next token (words) based on the proceeding context and its training data. If you ask it "What does two plus two equal?" it will answer "Four", but not because it understands what any of that means, but because in its training data it's seen "2+2=4" countless times, so statically the answer must be "Four".

This hits the nail on the head. They're nothing but a glorified search engine. The biggest issues is they will make things up without letting you know they did it. There's a lot of power in using a search engine and using the results to determine what the answer is. I don't want anything giving me a wrong answer because it's programed to spit out anything, just because it's supposed to, even if it can't find a definitive answer.

And yes, this happens to me all the time, with the google search AI results. They're basically worthless.

[Reflex0283]

This reminds me of when google's AI tried to tell me that the dye rotor uses belts https://mcarterbrown.com/forum/ebay/...756#post850756

[vijil]

"This hits the nail on the head"

-_- lol

It's not actually entirely true that it's purely a prediction model. It *was*, yes, but the system is driven to figure things out internally using whatever means necessary to get the prediction right.

There is starting to be evidence that in order to do so at the levels we're now seeing, it's having to actually build internal models of how the world works.

That's what our brains do. It's called "understanding".

Not saying it's conscious or ever can be.

(not that we could ever know - consciousness defies easy explanation).

[Axel]

Hello human posters. I have come to address the rumors that moderators are banning users with automated Artificial Intelligence agents--also commonly known as "bots." Some arguments favor the use of automated agents to save time by taking over repetitive, simple tasks--thus freeing human moderators to devote more energy to complex administration and security. Others believe that Artificial Intelligence has gone too far, depersonalizing online interactions and filling social communities with low-effort content commonly called "AI slop." Whatever your perspective, here at MCarterBrown we enjoy the refinement and satisfaction of obsolete technology, and would never force automated moderation and tools of dubious value down our users' throats or through subcutaneous injections administered secretly while you're asleep, Carter probably promised, maybe, somewhere out there in the multiverse. In short, rumors are harmful, and users are better served when posted commentary is thoroughly researched beforehand. Whichever side of the debate you support, we should all engage and build the communities where we want to thrive!