Whenever I feel in text, image, video AI is used with low effort then I just skip those. And Also sometimes I just skip High effort AI only post

Original

Now personally whenever see somethings then I prefer humans mostly compared to AI.

I personally use AI lot (10~30x compared to others) but I mostly use for learnings, fact checking, understanding etc.

Even I use AI for coding then I don't do like "just build browser, no mistakes..." etc. I use AI and then I invest same or even more time to refine, refactor and craft my things personally.

I also personally hate those who are great knowledagble persons in another domain and they share there point of view (to influence & sale) related to AI even without knowing how exactly AI works.

And today world the term AI we and I personally saying is not actually AI it's machine learning which just predict next token (not as simple as I said, I agree it's complex, but it's not an AI) I personally feel even large corps/tech knows it's not the AI but saying AI helping them that's why they are saying it. And also with the current structure (LLM) I personally feel we'll never ever achieve AGI with LLM.

For AGI we need to have completely different or atleast big enough modification in current way. I see LLM's no more greater then a programming (like python/js) function and this function have lot of logics inside it and it takes many parameter from it, and then the function return as value with pre-defined rules (pre-trained) and even we see like web crawling, real time then it's like in function passing additional awarenewss input, parameters etc.)

Someone will say but we can fine tune it. Yes we can but it's like copy pasting that same function and changing some logics for specific use case to recude future cost, increase speed and accuracy etc.

LLM's itself are really really really (3 times) great thing as of we've invented. Even my this post itself is used to fix grammers and sentences.

---

AI used to fix grammer (I know it's grammar)

Personally, whenever I see something, I tend to prefer humans over AI.

I use AI a lot—probably 10 to 30 times more than others—but mostly for learning, fact-checking, and understanding things better.
Even when I use AI for coding, I don’t just say, “Build this browser, no mistakes.” I use AI as support, then invest the same or even more time refining, refactoring, and crafting the final result myself.

I also dislike when highly knowledgeable people from other fields share their opinions about AI—often to influence or sell something—without truly understanding how AI works.

In today’s world, what we call “AI” isn’t actually AI; it’s machine learning that predicts the next token. It’s complex, yes, but it’s still not true artificial intelligence.
I personally believe even large tech companies know this but still say “AI helps them” for convenience or marketing.

With the current structure (LLMs), I don’t believe we’ll ever achieve AGI.
To reach AGI, we’ll need a completely different—or at least significantly modified—approach.

I see LLMs as no more than a highly advanced programming function (like in Python or JavaScript) containing many layers of logic. It takes multiple parameters as input and returns an output based on pre-defined, pre-trained rules.
When we see things like web crawling or real-time capabilities, that’s just passing additional awareness or parameters into the function.

Some might say, “But we can fine-tune it.” Sure—but that’s like copying the same function, tweaking its logic for a specific use case to reduce future costs and improve speed and accuracy.

LLMs themselves are incredibly—really, really, really—great inventions.
Even this post of mine was refined using AI to fix grammar and structure.