So you do understand the problem then.

Would it then be fair in your view to reframe the specific problem you have into the two following components?

1. This is the worst AI will ever be at being manageable by people. It will continue to improve until it's better, just like what happened with everything where AI is already better.
2. You can manage AI current gen AI with similar methods you'd need in managing your average "yes saar, of course saar, I'll go do what you say right away saar" Indian developer stereotype.

Notably, once you accept the second one, you quickly realize that you can use ControlNet style methodology of "just use a specialized AI to curate your inputs into your preferred task specific model". And for even better results, you can add model alloying into this specialized AI, so it can utilize the best way to handle the sycophantic worker. "Have a different worker check entirety of his work to see where the failures lie and fix them".

I can tell you never had to do managerial work, as you're unaware that one of the most common stereotypes of a worker. The guy who will say "yes boss" no matter what is asked of him, and you'll find out you asked too much of him only when he fails to do the task correctly and this failure is reported on. Often by someone else.

This is even worse with people that come from Indian culture, where "yes boss" is the expected answer regardless of how impossible the ask is.

What you're describing is fundamentally a managerial skill set.

A lot of software developers struggle with those, and quite a few are borderline incapable of it. That's going to be increasingly a problem, unless we manage to get AI trained on individual preferences, and correcting their responses into proper AI prompts. I.e. narrow model that AI incapable worker can interact with you generate a prompt for the major model that will do the actual work.

Ironically to get your computer to do what you want to, you need to train on how to use it correctly.

Prompting correctly is a part of using it correctly.

Is it worker's fault when manager fails to explain the task correctly?

This is the part that most of the "I can't make AI work" crowd miss. When you give AI instructions, it's like giving instructions to a human worker.

You need to understand its strengths and its weaknesses at least to a reasonable degree, and you need to be careful delineating what task entails, what it doesn't entail, what's a priority and what is of low relevance.

A lot of very good experts at their specific field make for horrible managers because they don't know how to explain aforementioned things. They only know how to do it themselves.

This is the most common point of failure both when leading people and when prompting AI in my experience. Essentially all those leadership skills? They matter a lot now, even for mere subject matter experts, because prompting AI is leadership just as ordering a person to perform a set of tasks with a specific goal is leadership.

Before most subject matter experts didn't need any meaningful leadership skills. They just needed to do the things related to subject they're expert in, and leadership was handled by people with a different skill set (less deep and more broad expertise coupled with at least some leadership skills).

Just remember ... it was almost impossible to get cellphone signal on planes during 9/11

Interested on Games Nexus take on this. I'm pretty sure it will be in their hardware news segment. I personally rarely use my Windows box anymore. My hacked PS4 Pro is my main gaming system now. My PS5 is sitting disconnected from the Internet until a 10.40 firmware hack is discovered. Fuck both nVidia and Sony.

This is a training problem.

Not of AI mind you, but of human doing the prompting. When AI fails this badly, it means prompt has poor instructions.

This is fixed by training people to prompt better. It's similar to teaching people how to be good managers. You need knowledge in subject matter, but your primary expertise should in be identifying, segmenting and communicating each individual part of the task with high precision.

This is not easy, and it's not the same skill set as that of typical coder.

I don't think theses LLMs are like any previous machine. You use a metal lathe or a cotton gin, you get reproducible results; near perfect results if you maintain the machine correctly and it doesn't have mechanical failure. LLMs for coding are not good deterministic machines. Tiny variations to a prompt or slight word ordering can dramatically change the output. Most models silently write away errors or do blank exception checking, making something seem like it works. It might be getting 50,000 records, but you're not seeing the 2,000 edge cases it's silently dropping. No one knows because no one reviews 10,000 lines of slop.

These machines are fundamentally worse than previous industrial revolutions. College students are booing people at graduations going on about AI because they've used it and the good students know just how wrong it gets things, sometimes 1/8 to 1/4 of the time! One day a bridge will collapse because a human check will be replaced by some shit engineering model and it will totally pass a design that has incorrect tensile strength. Than 4 ~ 10 people die and maybe, just maybe, we'll realize how dangerous these machines are ... NOT because they're intelligent game changes, but because they are next word predictive guessers with no intent and no real intelligence.

The machines won't kill us. Our trust in them will.

The entire economy is currently in a crashed state. Money printing went out of control, home compute is becoming unaffordable, no one wants any of these AI data centers, none of your OpenAI/Anthropic subscriptions even remotely cover the runtime cost of compute, much less the sunk cost of the infrastructure. It's all been run into the ground and we're all going through the motions like everything is fine. There was a term coined for continuing on in a collapse during the Soviet era: Hypernormalization.

We're distracted by irrelevant social issues, or people screaming about Orange Man (when really every single president in our lifetimes has been pushing towards this, unless you were alive during JFK, and we all know what really happened to him).

The data centers will get built, not by capitol raised through supply and demand, but because the governments of the world need the surveillance infrastructure to impose authoritarian, technocratic rule. That's the reason for "child safety" laws that are really digital identification laws designed to remove any remaining privacy anyone has.

Universal basic income only makes sense if there is zero resource scarcity. We don't have the dilithium crystals and replicators of Gene Rodenberry's universe. The Weighted Random Word/Code Generators are nowhere close to the level of Commander Data, nor are they "intelligent." They cannot reason, they just do a very fancy and compute intense predictive output. It's impressive, but transformers and attention blocks will never be an "AGI."

The technocratic class wants programmable money so you're only allowed to spend FOOD-COIN on x products and HOUSE-COIN on housing. Extreme control that can cut off anyone from their money should they say or think the wrong thing.

Universal Basic Income creates a permanent class tied to government gibs. It will be nothing like Star Trek and a whole lot more like The Expanse.

90%+ of "UFOs" are likely covert air force operations. The documentary Mirage Men interviews a CIA operative who intentionally drove a former Lt Col turned military contractor to a mental institution, in order to continue a narrative to cover up military operations.

It's hatred horse breeders and horse messengers had for cars and their drivers.

Notably, this is the exact reason why they had to immediately start having exceptions in the law.

In this case they started with farmers, and you'll likely have problems across insurance industry, which will require way more exceptions added on later.

Do meds make you stop thinking or analyzing things critically? Most of the psycho meds make people dumb as shit today. Here I have a whole writeup:

https://battlepenguin.com/poli...

There are sources for videos, analysis of the "bullet photos" and lots of other details. My arguments are sound. They might not be something you agree with because you believe everything you see on the Google and/or TV without question, but they are not an indication I need to "take my meds." Or have you forgotten how much the Epstein files have shown just how many "conspiracies" are true?

Operation Ajax. Operation Paperclip. Operation Mockingbird. COINTELPRO. All these are real, actual conspiracies against innocent people that took place under various presidents.

Keep taking your meds and go numb.

People don't take long haul trips in Europe. If they do, they'll often take a train and rent a normal car at the destination if they need one. America, Australia, etc. are really spread out. Every person I know who tried to do a long haul trip (like to the beach) in an EV said "never again." Especially in high vacation season when charging stations can have queues of 10~15 vehicles. If you wanted to have America on EVs for all trips, you would need charging stations the size of a Buc-ees everywhere, and they'd be mostly empty except for Thanksgiving, Memorial Day, etc.

Also, if you are dumb enough to do a long haul and find yourself in a 10 car queue for the charger .. go to an RV campground and rent a space instead. Keep some high voltage adapters and plug into the RV electric ports to charge.

EVs only work if you live in a city and never leave it. If you're in NYC, Chicago, you also need a garage (probably a space heater too) because you're not going to be able to charge those things at most charging stations when it's -10F outside. Most Americans can only have EVs as a secondary vehicle. The primary will always need to be gas or hybrid.