I did not intend to spend Christmas morning correcting this but I also did not expect people to spend Christmas Eve posting rape case court docs for clicks by pretending they're new.

Show respect for the victims. They have been presenting evidence for decades. Both parties ignored or smeared them.

It's true that Garland did nothing about Trump and Epstein while being completely aware of the crimes. The crimes were public information! Garland was installed to protect the DOJ, which protects Trump and Epstein. It's explained in this free 2023 article. sarahkendzior.substack.com/p/servants-o...

So don't use it for summaries. You've correctly identified that people use tools incorrectly sometimes.

"literally everyone" is a laughable statement that is not grounded in reality.

We haven't scratched the surface of use cases yet.

Go behind the scenes with the Jane Goodall Institute as it shapes the future of research and responsible innovation with AI - Ars Technica share.google/fJxfp6X85suZ...

Go behind the scenes with the Jane Goodall Institute as it shapes the future of research and responsible innovation with AI - Ars Technica share.google/fJxfp6X85suZ...

Not useless.

I think you give this person too much credit. But that is a kind and generous theory.

Math classes in high school incorporate graphing calculators into the curriculum. They're required. Education had to change. AI will force the same changes elsewhere too.

Hallucination rate is very low, going from over 30% to leads than 5% in most frontier models in just a year, and if you ground it with RAG techniques it's near zero. I've never seen a hallucination in Notebook LM. All sources are directly cited.

If you already know math, why use a calculator? It will just erode your math skill. It's no different.

It's incredibly useful for looking up things, facts, in large documents. Try it with Notebook LM. Load in your car manual and ask things. Don't just parrot what you've heard.

Narrowing scope doesn't change the size of the model dummy. What is wrong with you that you need to be insulting?

Which I do. Not sure how that makes them useless. Quite the opposite.

no, my extensive use of them has convinced me. they're incredibly effective for programming.

"more limited than claimed" is begging a lot of questions, but in one year they've gone from barely functioning to essential tools.

Computer programming in clinical research. I know a number of pharmaceutical companies using them in a variety of capacities too.

You didn't educate yourself. It does not make them smaller, it merely focuses them on a specific use case, altering the internal node weights.