FaceDeer

It’s not what more you should have said, but what less. It’s the “people were worried about nuclear armageddon” thing that’s the problem here. You’re making it look like the concerns about Y2K were overblown and silly.

You’re focusing on the extreme unrealistic end of what people were worried about with Y2K, but the realistic range of concerns got really high up there too. There were realistic concerns about national power grids going offline and not being easily fixable, for example.

The huge amount of work and worry that went into Y2K was entirely justified, and trying to blow it off as “people were worried about nuclear armageddon, weren’t they silly” is misrepresenting the seriousness of the situation.

Not necessarily. Curation can also be done by AIs, at least in part.

As a concrete example, NVIDIA’s Nemotron-4 is a system specifically intended for generating “synthetic” training data for other LLMs. It consists of two separate LLMs; Nemotron-4 Instruct, which generates text, and Nemotron-4 Reward, which evaluates the outputs of Instruct to determine whether they’re good to train on.

Humans can still be in that loop, but they don’t necessarily have to be. And the AI can help them in that role so that it’s not necessarily a huge task.

It means that even if AI is having more environmental impact right now, there’s no reason to say “you can’t improve it that much.” Maybe you can improve it. As I said previously, a lot of research is being done on exactly that - methods to train and run AIs much more cheaply than it has so far. I see developments along those lines being discussed all the time in AI forums such as /r/localllama.

Much like with blockchains, though, it’s really popular to hate AI and “they waste enormous amounts of electricity” is an easy way to justify that. So news of such developments doesn’t spread easily.

Funny you should mention blockchains. Ethereum, the second-largest blockchain after Bitcoin, switched from proof-of-work to a proof-of-stake validation system two and a half years ago. That cut its energy use by 99.95%. The “blockchains are inherently a huge waste of energy” narrative is just firmly lodged in the popular view of them now, though, despite it being long proven false.

A lot of work has been going into making AIs more energy efficient, both in training and in inference stages. Electricity costs money, so obviously everyone’s interested in more efficient AIs. That makes them more profitable.

The term “model collapse” gets brought up frequently to describe this, but it’s commonly very misunderstood. There actually isn’t a fundamental problem with training an AI on data that includes other AI outputs, as long as the training data is well curated to maintain its quality. That needs to be done with non-AI-generated training data already anyway so it’s not really extra effort. The research paper that popularized the term “model collapse” used an unrealistically simplistic approach, it just recycled all of an AI’s output into the training set for subsequent generations of AI without any quality control or additional training data mixed in.

Kids these days don’t learn cursive writing, it’s destroying their literacy!

Apparently, I am. People actually want this

Thank you for recognizing this. It gets quite frustrating in threads like these about new AI tools being deployed when people declare “nobody wants this!” And I try to explain that there are actually people that do want it. I find many AI tools to be quite handy.

I tend to get vigorously downvoted at that point, as if that would make the demand “go away” somehow. But sticking heads in sand doesn’t accomplish anything except to make people increasingly out of touch.

Based on your upvote/downvote ratio it looks like basic economics is not very popular.