Thats true for grok and other models. You need to generate endlessly and cut endlessly and stitch them up together.

but for seedance you pretty much get stellar output which is extremely consistent in a single shot. Like there is a very good chance you will like what you see.

Getting better at chess, the Rubik's Cube, etc. is a skill. The more you practice and the more passionate you are, the better you get at it. This is just like learning to play the piano or learning to swim. There is natural aptitude as well but its not as simple as being super smart and being able to strategize well using intelligence. That's just the narrative that seems to make sense. Being good at chess isn't evidence that someone is smart in general.

I do. I take the $500k from you and then pay $450k to a bunch of employees who put in 40 hours a week. I take home $50k. I earn $50k.

Now, you like the home so much you tell all your friends about it.

I hire more employees and I take $50k profit per house but I build 20k houses and make $1B. I earn $1B.

Market forces decide how much you "earn". There are plenty of blue collar workers who work 3 jobs and live paycheck to paycheck. There are SWEs who work 20 hours a week and make the bank.

At that time the predominant view was that LLMs were nothing but stochastic parrots, that they would plateau, and that hallucinations couldn't be fixed.

At this point I doubt there are any AI sceptics left. That ship has long sailed. The only thing that matters is whether the estimates are accurate, and AI can improve on that too.

Even humans only estimate based on neurons firing in prior patterns.

I think what you are getting wrong is thinking that the reader cares about your effort. The reader doesn't care about your effort. It doesn't matter if it took you 12 seconds or 5 days to write a piece of content.

The key thing is people reading the entirety of it. If it is AI slop, I just automatically skim to the end and nothing registers in my head. The combination of em dashes and the sentence structure just makes my mind tune it out.

So, your thesis is correct. If you put in the custom visualization and put in the effort, folks will read it. But not because they think you put in the effort. They don't care. But because right now AI produces generic fluff that's overly perfectly correct. That's why I skip most LinkedIn posts as well. Like, I personally don't care if it's AI or not. But mentally, I just automatically discount and skip it. So, your effort basically interrupts that automatic pattern recognition.

The main reason being scientific consensus can lag reality significantly, especially when career incentives discourage dissent. The history of science includes many cases where consensus was wrong and critics were marginalized rather than engaged.

Deference to science as an authority is the opposite.

Feynman has a quote on this:

"Science is the belief in the ignorance of experts. When someone says, 'Science teaches such and such,' he is using the word incorrectly. Science doesn't teach anything; experience teaches it. If they say to you, 'Science has shown such and such,' you might ask, 'How does science show it? How did the scientists find out? How? What? Where?' It should not be 'science has shown' but 'this experiment, this effect, has shown.' And you have as much right as anyone else, upon hearing about the experiments — but be patient and listen to all the evidence — to judge whether a sensible conclusion has been arrived at."

It is the same thing in sports as well. There will only ever be one Michael Jordan one Lionel Messi one Tiger Woods one Magnus Carlsen. And they are paid a lot because they are worth it.

>> Meta seem to be spending so much so they don't later have to fight a war against an external Facebook-as-chatbot style competitor

Meta moved on from facebook a while back.It has been years since I last logged into facebook and hardly anybody I know actually post anything there. Its a relic of the past.

Basically you give context to the AI models. Like if you want to build a social network you give the AI all the documentation around it so it can follow the best practices and guidelines etc.