How to pronounce “llm” in English

LLMs are a pretty good example of it.

LLMs, uh, large language models, generative AI, machines that seem to many of us to be intelligent, some say super intelligent, and more than that seem compassionate, seem empathic.

non-LLM generated systems of the past.

another LLM."

and LLMs, unregulated, and we will have a democracy, it's absurd.

- and LLMs are telling you where.

It looks like the LLM is not very good at this particular task, and he couldn't.

So you use the LLM to basically enrich the experience, maybe add more context.

If I ask the LLM, "Who won the soccer World Cup in 1998?"

of most LLM benchmark leaderboards.

これらのLLMはど れほど重要ですか？

Even building LLMs-- very high capex.

flexible adaptive concepts LLMs, uh, do exactly the opposite.

It means that if LLMs are learning human behavior through all these predictions, there might be a limit to which that approach can understand the underlying human

It’s just LLMs all the way down.

Something I noticed with LLMs-- I've been using Gemini, ChatGPT for a while.

Where do you see LLMs fitting into the questions you discuss in the book and course?

I put it in learn LLM, I made a podcast, I made flashcards, I made and she's like, got her whole incredible.

And maybe we leverage the LLM layer.

We learned that to an LLM, a sequence of jobs is a sequence of words.

- So that's on the LLM programming side, refactoring.

Have you ever used an LLM?

So you can think of in an LLM, maybe, all of the different data that's collected is integrated and associated according to meaning, right?

There was no tool up until LLMs that could allow you to put in what you need.

I think the reason that LLMs are so successful is because they allow for this kind of growth.

We built our own models before LLMs came in, so they can't possibly overlap that way.

It's not clear to me that LLMs or any current approach to understanding AI is positioned to solve that problem.

But people are using all these LLMs and different AI tools to abstract away that part of their life.

So when a company has to build its own LLM, even going beyond fine-tuning and pretraining-- there was a graph that Ravi had.

So if you think about the core logic of an LLM, it's based on the following problem, which is I have a sentence like, uh,

- Honestly, building an LLM from scratch is a lot of fun.

until it finds one that works to convince the other LLM to do something bad.

And here, with an LLM, you just say, give me the code for this.

So if you're putting that task off to an LLM, then you're not getting it clear, and you're not really understanding what you're writing about.

It's not in computable form until we had LLMs in generative AI.

And yet, most LLMs don't give you a place to write beyond prompting.

And these people tend to stick around because LLMs have so much to offer, even as you grow.

And because we're not using-- --we're not using common LLMs, you can't tell.

And you have a lot of people doing that with LLMs, with text-to-image models because they're so inflexible.

Instead of crunching numbers, LLMs suck in and spit out text.

There may be A fundamental limit in the ability for LLMs to understand that kind of transitional behavior Right

And since everyone is using social media and now LLMs for the past 10 or 20 years, we have to understand that those

It has Vera Rubin and NVLink 72 to run the LLMs.

be just a romantic human side of me saying that LLMs won't be able to capture that, maybe desperately holding on for hope.

Others use their algorithms to generate prompts to help train LLMS more efficiently than humans ever could.

And this is very similar to the kind of information that LLMs and other generative AI tools learn, is this kind of knowledge about the world.

and the space requirements for something like Google's latest LLMs.

I will show you how LLM calculates and completes a single sentence.

And then, I work with an LLM or with something else to iterate and make it better.