Greek Islands and Beyoncé

If you enjoy these newsletters and think a colleague or a friend should read them, please get them to sign up on this page . 📝

If it is your first time here or you want to binge-read past letters, here are all. 📰

If you found this newsletter interesting, I would love it if you could share it🔗!

👉 These are the crazy things I learned last week.

1/ Beyoncé

You know you are BIG when one of your concerts is a reason for a whole country to adjust its inflation. Read here how Beyoncé's concert managed to increase inflation in Sweden. Mind-blowing the power that some brands/creators hold.

2/ Paros

This is the story of how I decided to change my tickets just to run a 5k on a Greek island. 🏃🏻‍♂️

While I was vacationing in Paros (one of the amazing Greek islands you should visit), I saw that a day AFTER I was heading back to Athens, a 5k run was taking place.

As I'm trying to experience more  life , I changed tickets, booked a new hotel, and signed up. It's 5k; how difficult can it be? 🙂

Breaking news: Islands CAN have tons of hills.

But that's ok. It was fun! Until we had to get a boat to go back to Athens in THREE HOURS, and then our Airbnb was crap, and we had to relocate to a hotel at 11 at night.

But at least I got everything on camera 😉

3/ Cyclades

Ancient Greeks are THE BEST Storytellers. Here is why.

Leto was pregnant with Apollo and Artemis, and she was trying to hide from Ira because the father was Zeus, and well ... Ira was his wife. While she was trying to hide, right in the middle of the sea, a "traveling" rock submerged the island of Delos - which translates to "declaration"- to give a place for Apollo and Artemis to be born.

For the newborns to be protected, some islands circled Delos - this island formation is called Cyclades (translating to encircling).

Isn't it amazing that ancient Greeks didn't just see some rocks in the sea but a whole story? A story that was part of a bigger one! A story where eternal myths were created.

Storytelling is not about "interesting" stuff. It is about making "boring" stuff sound interesting, like some rocks in the water.

4/ GPT Engineer

Are you ready to have your mind blown away?

GPT-Engineer is the coding partner you've always wanted. And it's wild!

Working with GPT Engineer is very simple. You specify what you want to build with a prompt, the AI asks for clarification, and then the agent makes the entire codebase.

It's completely free and open-source, and it's already at nearly 35,000 stars on GitHub 🤯

5/ Open AI Function Calling

The AI space runs with incredible speed. 🚀

The latest release of Open AI allows developers to call functions directly from a language model.

Why is this important?

Well, if you have tried to ask for specific information from ChatGPT, for example, the weather in the city, you have been kicked with the usual response.

I apologize, but as an AI language model, I don't have real-time data or the ability to browse the internet. Therefore, I can't provide you with the current weather conditions in Boston or any other location. However, you can easily find up-to-date weather information for Boston by checking a reliable weather website or using a weather app on your smartphone.

Assuming that you had a function that gets the water based on the city, you could try to make the LLM call this function, but it would be challenging due to the non-deterministic output.

You'll probably spend a lot of time in prompt engineering, testing, and error handling, and it's not sure if you'll make it work.

Now the API allows developers to provide "function signatures" to the model, and the model can produce the arguments for your function signature.

The idea is simple:

1. You write a function with a signature
2. You define this signature in the OpenAI call
3. You get back the parameters from the model and the context
4. You parse the parameters and use them as real arguments.

For example:

input = "What is the weather in Boston?"

output = {"location": "Boston"}

return get_weather(**output)

Do you want to dive into technical details? Start from this Python Notebook about Function calling.

That's all for now. See you next week!

- Jon V (as in Victory) 🚀

- Fri 23 June 2023