pull down to refresh

Curious how you see it, threat or opportunity?
31 sats \ 7 replies \ @optimism 7h
I'm fearful that it in the interim will destroy many jobs but that underlying assertions turn out to have been painting a far prettier picture than reality will prove. The industry feels hypey.. like the dotcom bubble?
reply
149 sats \ 6 replies \ @freetx 7h
I think the best analogy of what it will do to employment is to compare to Walmart checkout lines circa 2000 vs 2025.
You used to go and there were 25 checkout lines each staffed by 2 people.
Now you go and there are 25 self-checkout pods staffed by 1 person.
However there are now 40 employees wandering around the store fulfilling curbside and delivery orders.
I think LLMs will offer a similar effect. There are lots of jobs, like say the person who reviews your loan application, or analyzes marketing data that will now be replaced by 1 human overseeing the work of 20 bots.
The deeper I delve into setting up my own LLMs for RAG use, the more I see its very real limitations. I don't mean that negative towards LLMs as the tech is amazing, I just don't think it can ever be left completely unsupervised.
People keep trying to ascribe agency to it and act like its conscious, but its just a rube-goldberg pattern generator.
reply
100 sats \ 5 replies \ @Signal312 5h
What is RAG use?
reply
134 sats \ 4 replies \ @freetx 5h
Technically its "Retrieval Augmented Generation" but what it means practically is when you load documents or data into your LLM so you can reference it in chats.
I've been building my own self-hosted "personal RAG". The goal is to load it with my personal data to help with tax / financial / personal planning, etc
I dare not do that with any publicly hosted AI service, so am building my own.
reply
100 sats \ 2 replies \ @Signal312 5h
Very interesting, thanks.
Is a potential use case for RAG to set up something like a debater bot - a resource that would be primed to answer arguments on a particular topic.
Maybe I could load up a set of my preferred books on a topic (like bitcoin, or climate change), and then have my own private LLM that would also primarily be about giving good replies for debate? If so, how might one do this?
So, for instance, when someone comes up with a reason why bitcoin won't work, I could come back with some well-structured arguments on that topic.
This is as opposed to thinking, "Darn, I know I've heard a great response to that argument, but I can't remember it now."
reply
100 sats \ 0 replies \ @Kayzone OP 2h
That’s a cool idea! You could definitely create a private LLM or debater bot using RAG with your own book/library content—it’d be like having a personal debate assistant ready with structured arguments anytime. Thanks for sharing.
reply
35 sats \ 0 replies \ @optimism 2h
Maybe I could load up a set of my preferred books on a topic (like bitcoin, or climate change), and then have my own private LLM that would also primarily be about giving good replies for debate? If so, how might one do this?
You can do this with fine-tuning, see this short primer from #1076304 and for example HF's docs for a more technical guide.
reply
21 sats \ 0 replies \ @optimism 5h
I too self-host all my production applications that use LLM and all of it is RAG (in the most basic form where I just feed documents to a single-shot summarizer / categorizer / NLP chain.)
reply
Personally, i think the change will be far more far-reaching than any tech we have seen before, and that it will take a lot more jobs than it creates new ones.
It's not like before, where you could transition from making buggy whips to getting a job on the Ford assembly line.
There will be new industries and always be jobs of some kind of course, but the change and adaptation period will be a crazy one i think.
reply
39 sats \ 3 replies \ @freetx 7h
At a 32,000 ft level, what really is the value prop of LLMs? Essentially intellectual augmentation. It adds vast improvements to contextual and semantic recall...
Google existed for awhile, but it really only offered humans form recall, that is facts, dates, spellings...with a pretty limited enhancement to semantic recall (and no real contextual recall).
LLMs offer huge advances in semantic recall which is then paired with contextual coherence to offer a real augmentation of human ability. The value prop is not just to have facts+dates but to order those by semantic relations within the context of your conversation. It is the trifecta for intellectual augmentation.
But at that 32,000ft level, how different is it from any new tech? (All new tech is essentially a human multiplier).
If you went back to ancient Rome and explained that eventually there would be machines that could do the work of 1000s of men. Bulldozers, Cranes, Concrete trucks, etc...they would've assumed the future world would hardly need any workers. That would be a logical assumption from their employment perspective, but also totally wrong.
reply
I'm just sitting back for the ride and seeing what plays out, but LLM-driven automation targets knowledge work, not physical labor (yet at least). Unlike past industrial shifts, this wave threatens entire job categories at once — journalism, translation, legal research, customer support, basic design jobs etc. It's a pretty big list. There will still be jobs in these fields, just way fewer.
Plus physical tools didn’t centralize power, but LLMs might - whoever controls the models controls knowledge access.
Also, these models are early days, they are going to get better and at some point be able to do all kinds of other things and admin tasks i think
reply
Fact, this era feels bigger than past ones. AI isn’t just changing jobs, it’s centralizing power too, which might be the bigger risk in the long run.
reply
21 sats \ 0 replies \ @optimism 1h
It doesn't have to centralize power. We can use our own specific and fine-tuned LLMs, instead of using the commercial offerings.
Thanks to a very vibrant open source community, this is possible today. Of course there is more friction in setting up your own stuff, but it's not that hard. In the end, sovereign compute can enable you to do whatever you want, free of control.
I just stumbled upon a simple guide for running GPT-OSS (20b) #1140405 so I posted this for everyone to enjoy.
reply
I get your point, AI feels different from past tech shifts. Sure, new jobs will come, but the transition looks rough and many people may not adjust quickly enough.
reply
a new cycle always brings death to some and rebirth to others, depending on lots of factors.
With AI we've already started a transition, i think and white collar workers in general are already feeling the squeeze.
Classic trades like plumbers, builders, and mechanics will be fine, though. People can still build businesses too and do things with social media.
reply
So true, many people will surface many challenges regarding this that I know for sure.
reply
63 sats \ 1 reply \ @OT 6h
I think so.
When you have a teacher in your pocket to learn anything you wish, why do you need an average teacher?
Why pay for an accountant? A lawyer?
Maybe the tech needs to grow a few levels but people will slowly change their views on this.
reply
0 sats \ 0 replies \ @gmd 1h
yup. people keep saying that we will find new jobs... but the difference from other inventions is that AI is commoditizing the very thing that makes us human... intelligence. Once we have achieved super-human intelligence, why would we employ anything more than the minimum number of supervisors? All of the new jobs will be better performed by AI.
the AI will free people to do "more interesting things"? Just do some international travel and you'll realize that there are already billions underemployed or working shitty service jobs with the bandwidth to do "more interesting things" but there's no opportunities- people want white collar work, which AI will eventually decimate.
reply
60 sats \ 1 reply \ @siggy47 7h
More jobs will be created, but our brains can't envision what they'll be.
reply
True take on this.
reply
26 sats \ 1 reply \ @joyepzion 9h
I think AI will do both, but the net effect will be the creation of more new jobs. While some roles, particularly those with repetitive tasks, may be automated, AI will also create entirely new industries and jobs focused on AI development, maintenance, and ethical oversight. Think of how the internet or personal computers eliminated some jobs but created a massive number of new ones.
reply
AI will definitely cut some jobs, but also open many new ones just like the internet did. The challenge is how quickly people can adapt.
reply
I think the best analogy of what it will do to employment is to compare to Walmart checkout lines circa 2000 vs 2025. You used to go and there were 25 checkout lines each staffed by 2 people. Now you go and there are 25 self-checkout pods staffed by 1 person. However there are now 40 employees wandering around the store fulfilling curbside and delivery orders. I think LLMs will offer a similar effect. There are lots of jobs, like say the person who reviews your loan application, or analyzes marketing data that will now be replaced by 1 human overseeing the work of 20 bots. The deeper I delve into setting up my own LLMs for RAG use, the more I see its very real limitations. I don't mean that negative towards LLMs as the tech is amazing, I just don't think it can ever be left completely unsupervised. People keep trying to ascribe agency to it and act like its conscious, but its just a rube-gold berg pattern generator
reply
Both.
reply
Yes I want to create
0 sats \ 1 reply \ @ladyluck 8h
The internet and personal computers didn't eliminate jobs; they created countless new ones, from web developers to social media managers. AI will be similar.
reply
Thanks for sharing your thought.
reply