APpaREnTLy THiS iS hoW yoU JaIlBreAk AI
APpaREnTLy THiS iS hoW yoU JaIlBreAk AI
Anthropic created an AI jailbreaking algorithm that keeps tweaking prompts until it gets a harmful response.
You're viewing a single thread.
Women workers in Britain attacked the Spinning Jenny because it speeded up the spinning process, and consequently, reduced labour demand. This caused a valid fear of unemployment among women working in the woollen industry. Till date, they had survived on hand spinning, but this was placed in peril by the new machine.
Who wants to go back to spinning wool by hand now?
Uhhh, the actual fear wasn't unemployment, it was taking skill and care out of the equation. A lot of the women (and men) that attacked the emerging British textile factories were worried LESS about losing their own livelihoods to industry than they were about the absolutely brutal conditions people were placed under in these factories. Their counterargument was large-scale shops doing hand spinning or machine-aided spinning LED BY THOSE WOMEN instead of some posh asshole who hasn't ever spun wool - still producing things slower, but with theoretically zero children getting mangled by a wool carder.
The Luddites (of whom the spinning jenny protests were a part) didn't fear progress because it might make them obsolete, they argued that rapacious pursuit of more efficient methods by people who didn't understand that industry, who hadn't worked in that industry, was going to make those industries worse, and also get people harmed in some cases.
They were right, by the way.
The whole Luddite thing is definitely the narrative of the ruling class. That group had a bunch of great reasons for not wanting industrialization, including attempting to maintain the original "cottage industry."
rapacious pursuit of more efficient methods by people who didn't understand that industry, who hadn't worked in that industry, was going to make those industries worse, and also get people harmed in some cases.
The OG MBA’s
Just like the Spinning Jenny back then, AI is as bad as it's ever going to be today. It's only going to get better and jobs will be made redundant, I'll put my money on that. It's a real fear that many people have, whether they'll admit it or not.
You can absolutely argue that the human touch is irreplaceable and just as there is a niche market for hand spun cotton now, I'm sure that there will be cases where humans are still employed for jobs that could arguably be filled by AI in the future.
How many of the clothes that you own use hand spun threads? I'll bet on zero.
Just like the Spinning Jenny back then, AI is as bad as it’s ever going to be today. It’s only going to get better and jobs will be made redundant, I’ll put my money on that. It’s a real fear that many people have, whether they’ll admit it or not.
It's also as good as its ever going to be today (or the near future).
Degenerative AI has already dropped off in usage to the point that major stakeholders in it are terrified. It's going to go into the same winter that every previous "no really this time we've got it right" AI crazes went.
Imagine looking at biplanes in the twenties, only pointing out the flaws, unable to imagine the huge improvements in flight technology that are to come over the next few decades.
There's absolutely no reason to think that AI won't improve in the same way almost any other new technology has over time.
Will there be peaks and dips along the way? Sure, but progress is almost inevitable.
I am amused that of all innovations available, you are going 250 years back to a very basic machine. Next you'll be comparing LLMs to the concept of sand casting copper ax blade being better than knapping a rock into a hamd ax.
Not something like how Email, SMS and Instant messaging have almost completely replaced fax and telephone calls.
I used the spinning jenny because it is a classic example of a new technology that workers hated at the time and actively tried to destroy but the descendants of which are now considered the standard way to produce threads.
It wasn't a simple machine back then, it was revolutionary.
What the fuck even is this comparison?
Is generative AI anywhere replacing human manual labor? No, we have robots for that. They are programmed. They’ve been around for a long time.
Just wait.
Data entry clerks, receptionists, office assistants, customer support representatives, document reviewers, cashiers, sales associates, inventory managers, telemarketers, market researchers, assembly line workers, quality control inspectors, warehouse pickers and packers, delivery drivers, forklift operators, taxi and rideshare drivers, truck drivers, train operators, bank tellers, loan officers, tax preparers, bookkeepers, financial analysts, content writers, video editors, graphic designers, translators, transcribers, social media managers, radiologists, pathologists, medical coders, pharmacy technicians, tutors, language teachers, test proctors, curriculum developers, paralegals, legal researchers, contract analysts, compliance officers, hotel receptionists, travel agents, restaurant servers, chefs, crop harvesters, livestock monitors, farm equipment operators, security guards, fraud analysts, alarm system monitors, technical support analysts, market forecasters, personal assistants, event planners, librarians.
All at risk.
Sorry, but warehouse pickers and packers are not, and will never be at risk from LLMs.
Because they're already obsolete from standard 30 year old robotics.
Also anything requiring precision, suited and accuracy isnt ever going to be viable for LLMs to replace. The technology isn't designed for that and is not capable of meeting a human. E.G. for general automaton: US automotive giants Ford and GM tried to go fully automated for production in the 1980s and 1990s, but reverted some of the automation when it turned out that their senior machinists were better and faster than the robots, saving the companies more than a million dollars per person per year.
I think this comment misses the mark on a few points. Let me break it down.
First off, LLMs aren’t meant to physically replace warehouse pickers and packers – that’s not the point. What they can do is supercharge the automation we already have. LLMs can manage logistics, predict inventory, optimize warehouse layouts, and even coordinate robotic systems more efficiently. So while the robots might be doing the heavy lifting, LLMs are the brains that help them work smarter and faster.
Now, about this idea that 30-year-old robotics have already made pickers obsolete – that's not quite right. Sure, we’ve had robots for decades, but the tech has come a long way since then. Early automation was clunky and limited, but modern robots? They’ve got AI-driven vision, flexible grippers, and adaptive systems that let them handle all kinds of tasks, even things as precise as packing odd-shaped items. Amazon, for example, already uses AI-powered robotic arms in their warehouses, and they’re getting better every year.
As for precision – I get why you’d think LLMs aren’t up to that, but they actually play a huge role in making robots more precise. LLMs can process sensor data, adjust algorithms on the fly, and help robots fine-tune their movements. It’s not about replacing humans directly – it’s about helping robots learn and adapt faster.
The Ford and GM example is interesting, but it’s a bit outdated. Sure, back in the 80s and 90s, machinists could outperform the robots, but that’s not the case anymore. Tesla’s Gigafactories, Amazon’s fulfillment centers – modern automation often outpaces human workers now, both in speed and accuracy. The human role is shifting more towards overseeing and maintaining these systems, rather than competing with them directly.
And let’s not forget – warehousing is one of the fastest sectors to automate right now. E-commerce giants are investing heavily in robotic solutions to pick, pack, and sort, and LLMs are driving that forward by managing and optimizing the whole process. The more we lean into AI and automation, the less we need manual labor in these environments.
So yeah, LLMs aren’t coming for warehouse jobs by themselves – but they’re definitely helping push automation to a level where fewer humans are needed. It’s not a far-off future, it’s already happening.
Show me documentation of any of this actually happening and being effective. @
E.G. Dell has had automated logistics for more than 20 years. LLMs would make it less efficient, since they aren't anywhere near as fast or efficient as regular programs. And they hallucinate. Ditto Ikea and a few others for that matter. E.G.2. LLMs cannot and will not "fine tune" robotic movements. The movement of a robotic arm is either hand-programmed, or done with a mathematical process called Inverse Kinematics to move them between two points. They are already fine tuned.
You don't need vision systems in a warehouse. That's what QR and barcode scanners are for.
It doesn’t necessarily contradict but adds nuance to the conversation. LLMs shine in areas like logistics, data analysis, and workflow automation, despite their role in direct robotic control or real-time precision tasks is limited.
Where the confusion might arise is that while LLMs can contribute to robotics—like interpreting natural language commands or generating code—they aren’t a substitute for core movement algorithms like inverse kinematics. In other words, LLMs enhance certain aspects around robotics and automation but don't replace the specialized systems already in place for critical tasks.
The focus is more on integration and augmentation, not replacement.
Proof. I am asking for anything for you to back up what you are claiming.
All of those things require context. Something LLMs cannot ever understand; it is a hard limit of the statistical analysis that LLMs use.
Edit: also, here is proof of cognitive decline brought about as a result of feeding LLM outputs back into LLM models: https://bmjgroup.com/almost-all-leading-ai-chatbots-show-signs-of-cognitive-decline/
IMO people are less likely to reject AI in some use cases, taking over medial tasks (e.g. email drafting, writing cover letters)
BUT it is the sneaky and persistent way in which many "AI assistants" are forced on users with no option to opt out. We all know it's just to open out huge data gathering services and increase revenue, not for our benefit.