Pennomi @ pennomi @lemmy.world

Posts

1

Comments

1,661

Joined

2 yr. ago

Giving it the image as part of the context helps. But honestly for a description that short it takes longer to have an AI do it than to just do it yourself

Yeah that’s a good point. I probably should have said evangelicals, and noted that even then it’s a fairly fringe belief.

Most Christians believe that only devout Christians will be raptured, therefore clearly excluding people of other faiths.

Until someone comes up with a very compelling anti-drone countermeasure, drones are basically the ideal weapon for a battlefield. Cheap, mobile, can loiter, good recon, etc.

The sad part is that it doesn’t have to fool the people that are paying attention. If it even fools the foolish, humanity is in for a bumpy ride.

It’s interesting how quickly people learned to detect the patterns of AI images. Like, I can nearly always tell which platform it was generated on because they all have specific biases.

How is what I said AI apologism? I clearly expressed concern that the technology isn’t reliable or cheap.

Yes, it will take a substantial change in LLM architecture to make them viable for general use. Like you said, a couple orders of magnitude improvement in training would be the biggest help. Though I’m not sure that’s even possible.

Or it will become foundational tech like the Internet. People said that was a fad too.

It all depends on whether or not it becomes reliable and cheap.

They all died in a freak longboat accident. The Vikings responsible were never apprehended.

The more likely theory is that the billions of dollars spent on propaganda actually worked. Marketing is far more effective than you’d think it should be… The average person is ridiculously gullible.

It’s not just democracy, but information quality in general.

This is what we get for seeking profits more than truth.

Except ConcernedApe, apparently.

You have to understand, programmers are the laziest bunch of people out there. We literally make our living telling the computer to do shit for us.

Now you’re telling me that I can tell the computer to tell the computer what to do? Great!

(/s if you can’t tell)

I’ve bought it at least 5 times; sent it to a lot of family members.

It seems clear that during training they weight Musk’s tweets (and those of accounts he likes) more heavily than everyone else’s.

What is the cutoff point where you can consider something to be one species or another?

Not an easy question to answer.

Agreed, drug development is a very good use of AI.

That’s the purpose of the alignment step in training, teaching it what things they learned are morally repugnant.

Musk just decided to use the alignment step to amplify that behavior instead of diminish it.