> manipulated imagery

And we thought iPhone camera videos were bad... (they were (and are) though)

I never understood what people are trying to say with comments like that.

- You're making unsubstantiated claim

- personally targeting someone you don't even know

- in order to celebrate presumed success of a mass fraud?

Effort makes a great deal of difference for me. The effort itself, the fact that it's there.

I am willing to suspend disbelief for Terminator 1, even if it is clear, that it's a head of the doll in shot.

But it is insulting to feed slop to your audience; it shows you didn't even try.

I have actually seen one slop-video, that I kinda enjoyed - it was obvious, that a great effort was put in a script and details as much as it was obvious it isn't being passed for the real thing.

We also need to take into account, that CGI only consumes energy when the actual creation of particular video happens.

"AI" consumes energy before user even started (during training).

That is on top of comparison for each particular case.

Right idea, but the application is incorrect.

Model training is similar to the creation of the cgi for the movie. Both happen before anyone consumes the output, and represent the up front cost for the producer.

Both a movie and a language model can cost tens or hundreds of dollars to produce.

In both cases additional infrastructure is needed for efficient usage: movie theaters or streaming platforms for movies, and data centers with the GPUs for LLMs. This is also upfront (capex) costs.

At consumption time, the movie requires some additional resources, per viewing, whether it's a movie theater or streaming. Likewise, an llm consumes some resources at inference time. These are opex. In both cases, the marginal cost for inference/consumption is quite low.

  > Model training is similar to the creation of the cgi for the movie. Both happen before anyone consumes the output

I did not say anything about consumption of the output. Maybe you misread what I wrote, it is about energy consumption.

  > Both a movie and a language model can cost

But we weren't comparing cost of the movie to cost of a language model

  > can cost tens or hundreds of dollars

But we weren't talking about dollars, we were talking about energy.

We're clearly exploring different questions.

And that energy costs money, both at the training/cgi stage and at the inference/consumption stage. It's not even an externality.

CGI renders do use a lot of electricity relative to playing back the movie for individual viewers. It's perfectly analogous.

  > CGI renders do use a lot of electricity relative to playing back the movie for individual viewers. It's perfectly analogous.

I've literally laughed at loud after reading this.

I can't believe you're stretching this in a good faith.

But if you are - well, you're certainly have a unique perspective.

This is incorrect.

Is it?

https://thewolfgroup.com/blog/executive-order-us-professiona...

Only certain services?

Only to certain entities.

It is fascinating, that you took that quote and, somehow, managed to arrive at the opposite conclusion, while presenting this quote as confirmation.