ChatGPT's Open source equivalent just dropped.

Good Morning AI Runners

Today, we have a few things we want to talk about.

• You chat: AI chatbot with citations

• Open source ChatGPT equivalent just dropped.

• Reinforcement learning, explained.

You Chat

You Chat is a new conversational AI that is capable of having open and broad conversations about a wide range of topics, including current events.

Unlike chatGPT which is still stuck in 2021 and thinks queen Elizabeth is still alive.

You Chat is also different in that it can provide citations for its answers. While You Chat is generally up-to-date and accurate, it is still in beta testing and may make mistakes.

The developers of You Chat explained that they hope that by providing citations, apps, and web links alongside the chat function, users will be able to easily verify the accuracy of the information provided.

Can’t explain how exciting it is to think of all the new and improved chat bots coming in 2023 alone.

Jk, we’ve seen enough of them already but what we want is a reliable one with accurate answers and citations. Seems to be closer than ever now.

Try it out here, It includes the added feature of the you search and app platform:

https://you.com

Sick domain.

Open source ChatGPT equivalent just dropped.

Phil wang or “Lucidrains” recently released the open source equivalent to ChatGPT. He implemented a framework to train PaLM (Google’s 540 billion parameter language model) using the same reinforcement learning strategy as OpenAI’s ChatGPT.

Can you believe it? It hasn’t been 2 months since OpenAI released ChatGPT and we already have an open source equivalent on our hands.

(You can scroll below to the PPPs where we explain what reinforcement learning actually means)

This is still very new so we haven’t seen what people can and will build with it.

Just know that the genie is out of the box and you can’t put him back in 🧞‍♂️

Keep an eye out for it and we’ll be sure to let you know if we see anything cool built with it in the future posts of RunTheAI.

Github repo:

Run The AI: PPPs

Pick up (learn):

To be quite honest with you, this section has been acting as our own little learning journal. And if I am being even more honest with you, then this newsletter is run by me (alone) but for some reason I can’t help but use “we” and “our” every now and then.

Anyways, today I want to explain an interesting topic in AI, Reinforcement Learning.

Reinforcement learning is a type of machine learning that involves an agent taking different actions in an environment, in order to make good decisions that maximize rewards.

Imagine you're playing a video game where you can earn points for doing certain things and lose points for doing other things. Reinforcement learning is like having a computer try to figure out the best way to play the game and earn the most points.

One way reinforcement learning is used is in self-driving cars. These cars need to make lots of decisions to drive safely and avoid accidents. Reinforcement learning algorithms can help the car learn the best way to make these decisions.

Overall, reinforcement learning is a way for systems to learn how to make good decisions and optimize their behavior in different situations.

Pilot (play):

https://you.com

Person:

Yann LeCun, besides having a cool name, has been credited for advancing convolutional neural networks. Beginning in the late 80's he proposed architecture for building neural networks that would help computers recognize images.

Pic of the day:

That's it from RuntheAI for today.

THANK YOU FOR READING AND SEE YOU TOMORROW, SUBSCRIBE TO STAY UPDATED!