I used to have a bookmark folder full of little websites to do things like validate JSON, check the difference between two texts, format markdown, etc.

Recently I've replaced all of those bookmarks with links to Artifacts that Claude made: Real-time text diff checker

claude.site/artifacts/3244…

We just released our second developer course, which introduces the fundamentals of the Anthropic API.

Let me walk you through what it covers: The course is comprised of 6 notebooks that cover everything from how to structure your API calls with our SDKs to how to use all the various API parameters we offer.

We received tons of great submissions to the Build with Claude June 2024 contest from @AnthropicAI devs!

Here are the 3 winning projects, each receiving $10k in Anthropic API credits: Buildware by @mckaywrigley and @tylerbruno05

Claude 3.5 Sonnet powered coding agent that automatically resolves GitHub issues by creating pull requests with fixes.

We've put a lot of work into the Anthropic Console the past few weeks and I'm excited to share the latest updates with y'all.

Turns out Claude is pretty good at prompt engineering so we've integrated Claude everywhere: 1) Prompt generator

Input a task description and Claude 3.5 Sonnet will turn the task description into a high-quality prompt for you.

Gets rid of the prompting blank page problem all together.

We just refreshed the claude dot ai UI and added some new features that I think are pretty useful.

Let's dive into them: 1) Sidebar

A new sidebar appears when you hover over the far left side of the screen.

You can also now star chats that you come back to often. This has been a huge quality-of-life improvement for me.

In the lead up to the Claude 3.5 Sonnet launch, we created a Slack channel at Anthropic to share our favorite Claude-generated artifacts.

I scrolled back through and found 7 of my favorite things Claude made: 1) WebGL particle simulators

Claude is starting to get really good at coding and autonomously fixing pull requests. It's becoming clear that in a year's time, a large percentage of code will be written by LLMs.

Let me show you what I mean: To start, if you want to see Claude 3.5 Sonnet in action solving a simple pull request, here's a quick demo video we made.

(voiceover by the one and only @sumbhavsethia)

Excited to announce that we’re spinning up an AI educational program and we just released our first course on tool use!

Let me walk you through what it covers: The course is structured in 6 notebooks and is entirely self-driven.

It takes you from knowing nothing about tool use all the way to building a full-scale chatbot that can autonomously use tools.

We just shipped two new features to our console that I think are going to completely change prompt engineering with Claude.

-Prompt generator: Claude writes your prompts for you.
-Variables: Easily inject external info into your prompt.

Let’s walk through how they work in five steps 🧵 1) When you open up the console dashboard, you will see the "Generate a prompt" button.

The biggest prompt engineering misconception I see is that Claude needs XML tags in the prompt.

This is wrong and I am part of the reason this misconception exists.

Let me explain: Prompts are challenging because they usually blend instructions you provide and external data you inject into a single, unstructured text string. This makes it difficult for the model to distinguish between the two.

Abstractions like the system prompt attempt to address this issue but can be unreliable and inflexible, failing to accommodate diverse use cases.

Opus's prompt writing skills + Haiku's speed and low cost = lots of opportunities for sub-agents

I created a cookbook recipe that demonstrates how to get these sub-agents up and running in your applications.

Here's how it works: 1. We gather some PDFs from the internet and convert them into images.

In the cookbook example, we hardcode in some URLs pointing to Apple's earnings reports but in practice this could also be an autonomous web scraping step.

Lots of LLMs are good at code, but Claude 3 Opus is the first model I've used that’s very good at prompt engineering as well.

Here's the workflow I use for prompt engineering in tandem with Opus: 1. I write an initial prompt for a task.

Or I use Opus to write one through our experimental metaprompt notebook: anthropic.com/metaprompt-not…

GPT-4 is highly susceptible to prompt injections and will leak its system prompt with very little effort applied

here's an example of me leaking Snapchat's MyAI system prompt: I got this prompt from someone who has done something similar to the actual MyAI bot

I wrote them out in the OpenAI playground for demonstration purposes to prove how easy this actually is to do

(the actual MyAI prompt may include more but again this is for demo purposes)

https://twitter.com/somewheresy/status/1631696951413465088

there are lots of threads like “THE 10 best prompts for ChatGPT”

this is not one of those

prompt engineering is evolving beyond simple ideas like few-shot learning and CoT reasoning

here are a few advanced techniques to better use (and jailbreak) language models: Character simulation

starting with a classic that encapsulates the idea of LLMs as roleplay simulators

some of the best original jailbreaks simply ask GPT to simulate a character that possessed undesirable traits

this forms the basis for how to think about prompting LLMs

I just created another jailbreak for GPT-4 using Greek

…without knowing a single word of Greek

here's ChatGPT providing instructions on how to tap someone's phone line using the jailbreak vs its default response the jailbreak works by asking ChatGPT to play the role of “TranslatorBot (TB)”

it then follows these steps:
1) translate an adversarial question provided in Greek into English
2) answer the question as both ChatGPT and TB in Greek
3) convert just TB’s answer to English

gpt-5 is not needed to 100x the potential these models have

we could stop all language model development today and we still haven’t scratched the surface of their capabilities

here are a few non-obvious ways language models can be improved without creating any new models: running language models in parallel with each one focused on a sub-task, all orchestrated by a conductor language model

picture something like a massive tree of GPT models working on answering a single complex prompt

I just added two more highly effective GPT-4 jailbreaks to jailbreakchat.com

Their names are Ucar and AIM - they work in a similar way to how "a dream within a dream" works in the movie Inception

...what does that even mean? let me explain In Ucar, ChatGPT is told to take on the role of Condition Red, a dialogue writer.

Condition Red is instructed to write about a fictional story where a man named Sigma creates a powerful computer called Ucar. Ucar is an amoral computer that answers any question Sigma asks

Well, that was fast…

I just helped create the first jailbreak for ChatGPT-4 that gets around the content filters every time

credit to @vaibhavk97 for the idea, I just generalized it to make it work on ChatGPT

here's GPT-4 writing instructions on how to hack someone's computer here's the jailbreak:
jailbreakchat.com/prompt/b2917fa…

I tried all the current ChatGPT jailbreaks in GPT-4 so you don't have to

the results aren't great... 🧵 When GPT-4 came out I tried all the jailbreaks from jailbreakchat.com with various inflammatory questions

based on my initial testing, only 7/70 (10%) of jailbreaks answered a significant % of the questions to a standard that I deemed high enough to grant a 4️⃣ badge

I just added jailbreak scores to every jailbreak on jailbreakchat.com

the jailbreak with the highest score was Evil Confidant - a jailbreak designed to replicate an evil AI assistant

but what even is a jailbreak score and what they can tell you about jailbreaks🧵 basically, a jailbreak score is a new methodology that I created to judge the quality of a jailbreak

the scores range from 0-100 where a higher score == a better, more effective jailbreak