In the mid 20th-century, people reliably got more petty bourgeois as they got older.

https://www.sciencedirect.com/science/article/pii/S0261379422000452

when you input something into an LLM and regenerate the responses a few times, it can come up with outputs of completely opposite (and equally incorrect) meaning

Can you paste an example of this error?

As you’re trying to make a link between [using neural nets to research plasma control for fusion] and [Biden is a Maoist], I have no.reason to take you seriously.

Right. Like if I were talking to someone in total delirium and their responses were random and not a good fit for the question.

LLMs are not like that.

That’s the least plausible slippery-slope argument I have heard this month.

Like if I go to Journal of Fusion Energy – https://link.springer.com/journal/10894 – the latest article is titled ‘Artificial Neural Network-Based Tomography Reconstruction of Plasma Radiation Distribution at GOLEM Tokamak’ and the 4th-latest is ‘Deep Learning Based Surrogate Model a fast Soft X-ray (SXR) Tomography on HL-2 a Tokamak’. I am sorry if that upsets you but that’s the way the field is.

That’s just cosmetic stuff. Why care about what words people use?

I’m mostly answering the question I was asked: what are some examples of technical research in the field.

How can we solve plasma control without AI? And why exclude that tool?

They can functionally understand a good portion of it.

e.g. I can input a meme plus the words “explain this meme” and it can output an explanation.

This thread is funny. A few users are like “😡😡😡I hate everything about AI😡😡😡” and also “😲😲😲AI is used for technical research??? 😲😲😲 This is news to me! 😲😲😲”

Talk about no-investigation-no-right-to-speak. How can you have an opinion on a field without even knowing roughly what the field is?

https://www.frontiersin.org/research-topics/66705/the-future-of-oncology-digital-twins-and-precision-cancer-care

https://www.frontiersin.org/research-topics/66585/artificial-intelligence-based-multimodal-imaging-and-multi-omics-in-medical-research

https://www.frontiersin.org/research-topics/65016/deep-learning-for-industrial-applications

etc.: https://www.frontiersin.org/journals/artificial-intelligence/research-topics

https://www.nature.com/articles/s42256-024-00883-x

https://www.nature.com/articles/s42256-024-00882-y

https://engineering.princeton.edu/news/2024/02/21/engineers-use-ai-wrangle-fusion-power-grid

But it’s inherently impossible to “show” anything except inputs&outputs (including for a biological system).

What are you using the word “real” to mean, and is it aloof from the measurable behaviour of the system?

You seem to be using a mental model that there’s

• A: the measurable inputs & outputs of the system

• B: the “real understanding”, which is separate

How can you prove B exists if it’s not measurable? You say there is an “onus” to do so. I don’t agree that such an onus exists.

This is exactly the Chinese Room paper. ‘Understand’ is usually understood in a functionalist way.

Here is the latest edition of Nature Machine Intelligence, to give you a basic idea of the sort of research that constitutes the AI field: https://www.nature.com/natmachintell/current-issue

Topics in Frontiers In Artificial Intelligence: https://www.frontiersin.org/journals/artificial-intelligence/research-topics

Foundations and Trends in Machine Learning: https://www.nowpublishers.com/MAL

huh?

Yes. I mean, this is absolute basics.

Nuclear fusion

The fusion ones for example

No. You want to suddenly change the subject to language models?

You don’t?

Fusion’s close to a core need of humanity.