bestcoder69

Llamafile: https://github.com/Mozilla-Ocho/llamafile

permitted to force RTO without any data != ought to force RTO without any data

I can already use multiple backends by writing different code. The value-add langchain would need to prove is whether i can get better results using their abstractions compared to me doing it manually. Every time I’ve looked at how langchain’s prompts are constructed, they went wayyy against LLM vendor guidance so I have doubts.

Also the downside of not being able to easily tweak prompts based on experiments (crucial!)

And not to mention the library doesn’t actually live up to this use case, and you immediately (IME) run into “you actually can’t use a _Chain with provider _ if you want to use their _ API”, so I ultimately did have to care about whats supposed to be abstracted over

They’ve been growth hacking the whole time pretty much, optimizing for virality. Eg integrating with every ai thing under the sun, so they could publish a seo-friendly “use gpt3 with someVecDb and lang chain” page, but for every permutation you can think. Easy for them to write since langchains abstractions are just unnecessary wrappers. They’ve also had meetups since very early on. The design seems to make langchain hard to remove since you’re no longer doing functional composition like you’d do in normal python - you’re combining Chains. You can’t insert your own log statements in between their calls so you have to onboard to langsmith for observability (their saas play). Now they have a DSL with their own binary operators :[

VC-backed, if you couldn’t guess already

openai cookbook! Instructor is a decent library that can help with the annoying parts without abstracting the whole api call - see it’s docs for RAG examples.

They released chat and non-chat (completion) versions of 3.5 at the same time so not really; the switch to chat model was orthogonal.

e: actually some of the pre-chatgpt models like code-davinci may have been considered part of the 3.5 series too

I find I can answer questions regardless of how nicely they’re phrased

What does it mean for either side to “have an objective”? I’ve seen a slew of footage of IDF up and down the ranks clearly state and commit to that objective. Is there some narrow definition of objective that helps you sleep at night that you’d like to share?

limewire teens were pioneers too

together.xyz, although they’re .ai now

Newt Gingrich proposed once that we make kids do this, but it was just gonna be for the poor kids.

Tesla could just negotiate instead of being antagonistic.

Prosecutors need snitches.

> This extreme scenario is unlikely to happen in the short term because of the high costs associated with additional AI servers and bottlenecks in the AI server supply chain, de Vries says.

If we suddenly start ignoring the costs, it could get very expensive, indeed.

Prompts:

- does this API design make sense?

- my team thinks __ is a good pattern, what are some cons?

- what’s a good name for a class that does x,y,z

- is this code snippet readable?

- write a bash script to automate this very tiny thing

- upload quickjs binary to the advanced data analysis model and ask it to micro benchmark a couple of approaches to the same thing

A cleaner policy would be to ban cars. Probably need a transition in between, but that's definitely the destination. In the year 3000 our asses better not be sitting in traffic in single-passenger vehicles w/ 3 seats empty still for some reason.

Maybe workers will strike for... more detailed information about each worker's individual profit margin?

And if I could breath underwater I wouldn't need a scuba tank.

He's analyzing what looks good in marketing copy.

Maybe if you have a model with a large context window, you stuff a document in the prompt as a prefix, then ask a bunch of different questions about the document?

Self-driving cars work really well so far. But we'll see, I guess! https://arstechnica.com/cars/2023/09/are-self-driving-cars-a...

My kids are too young, tech will advance too much to know. I'll report back in about a decade about my policy around how much they're allowed to meld with F̵̡͇͈̬͍͈͉͛͗͌̓̂̊̅̓̀͆͝Ṟ̵̣̳͎̽̈́͘İ̴̛̞̹͉̲͎̿̌̇̋͑͗̒̅̚͜͝͝Ȩ̷̻͖͓͍̙̲̮̰̹̖̥͙̤̺̏͒̃N̸̛̛͈͎̖̩̲͔̫̫̗̲̆̍̌̀̾̽͐́̈́̓̚͘D̵̡̡̯͙̦̎̈́͝.

text-davinci-002 used to make me so mad with how often it’d do that

How well does it work?

Not leaking!

When you see this kind of text you’re just in a weird state and it’s going to look like GPT was talking to someone, but it’s really just babbling with no purpose

e: Get GPT-4 to complete an empty prompt, then ask it what it was responding to! I just tried with Simonw’s llm CLI like so:

  llm -m 4 ‘’
  # it outputs a weird response
  llm -c ‘What question was that in response to?’

In my case an explanation of euthanasia, and my supposed question was “What is euthanasia?”. I did it again and it said there was no original question, so there’s some randomnes.

That’s not what’s happening. endoftext is OpenAI’s document splitting token, so when GPT encounters it, it treats it like the end of a chat. So the chat “ends” as far as the model is concerned but the app is still sampling tokens, so GPT just… does its best. The output is similar to when you hit “Generate” with no prompt in OIA playground’s text completion mode. There, you can imagine an invisible endoftext as your prompt.

No, you can easily reproduce this behavior yourself (I did)

I'm wondering if this was all in a single ChatGPT session where GPT-4 happened to answer the first question wrong. This could bias future outputs towards being wrong as well, and could explain the huge difference between OP's results and commenters, maybe?

That Twitter user is making up details and writing a fan-fic: https://twitter.com/8teAPi/status/1686217806298423299?s=20

So you won’t find any other sources, other than other people who have also been misled by @8teAPi.

Frustrating because it’s also chaotic, trying to keep the events straight.