I think learning from another human being live who can understand your confusingly articulated questions and who can adjust their answers for you, another human being, that's great.

I agree, at least in that we need more human contact, not less.

But I will say Gemini / Claude have both been surprisingly good at understanding my half-assed attempts to put a complicated question into a coherent sentence.

Try being an English speaker learning Mandarin from Claude

What obstacles do you tend to run into? I'd be inclined to use a Chinese model for that, since Claude was presumably RLHF'ed primarily by native Anglophones.

Have you tried a Chinese-trained LLM for learning Mandarin, and if so, did it help?

I think to learn a language you need to do as many modalities as you can: Speaking, listening, reading, singing, asking questions to native speakers, asking questions to LLMs, writing, flash cards, etc.

The obstacle with LLMs is that they are specifically going at their own uncanny valley pace. I do see how that might be not a problem for some prompts.

I haven't tried a Chinese-trained LLM for learning Mandarin, so I'm not sure which LLM is best..

>Too costly. Now its viable with AI to do it cheaply.

Where is the data that shows this sort of technology beats normal pedagogy?

From a study I read that came out last year, AI tutors are at best no better than a teacher, and whats worse is that AI tutor use caused many students to become dependent on them. These students could not recall information or perform nearly as well without them.

This is just another example of a long line of behavior from software engineers thinking they are god's gifted people, who can waltz into any industry and revolutionize it overnight with an application.

Please support /voice for claude!

thanks for the feedback. i'm working on making voice input more seamless.

I think it's easy to lie to consumers. Then, how much agency do they have?

I heard people say this before. I'm wondering, how do you instruct the LLM to generate the tests? Do you tell it the scenarios that would be covered, or do you just tell it to write tests for the code?

>cover all edge cases

That’s probably the extent of the prompt.

I have only one thing to say....

what

I thought it was known since a few years now that if you train models to NOT do certain things, then they start behaving in weird ways…

It seems like they run a classifier model before going to Fable (or falling back to Opus), so it should be fine

What if the organization who tries to verify sends a request on an app on the user’s iPhone (or whatever device can do the same), and the user scans their face with FaceID to produce a file send to the organization, which will then send that file to Apple to ask if the file represents the right person? I trust Apple so that works for me.

I’d say it’s possible to have creativity when you’re sitting as well. I like to think that’s it’s all about staying active. Reading, diarying, calling a friendind. All of that.

LLMs are like a search engine that autocompletes. It's a tool.

What negative consequences does being unelected have?

Maybe you can't 100% know what every layer "thinks", if you go through all the layers, you might see a cohesive "thinking" story. So, if there is any information you lose at layer N, you might learn some of it in layer N+1. The masking in the layers is not deterministic so the model can't really consistently lie throughout the layers. It doesn't chose what information we get to inspect. There might be a game of whack-a-mole, but you might get a general sentiment. I think the more layers there are, the more the model itself can hide very nuanced lies (But by that time we'd have a better mind-reading model).

However, I haven't read about it yet. I'm really excited to look into it!